diff --git a/mis_dro/constants.py b/mis_dro/constants.py index 040a4f1..4caf8a5 100644 --- a/mis_dro/constants.py +++ b/mis_dro/constants.py @@ -52,6 +52,40 @@ 3, ] +WASSERSTEIN_DRO_EPSILON_SET = [ + 0.001, + 0.005, + 0.01, + 0.05, + 0.1, + 0.5, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 12.5, + 15, + 17.5, + 20, + 22.5, + 25, + 27.5, + 30, + 35, + 40, + 45, + 50, + 55, + 60, +] + + SMALL_BAS_DRO_EPSILON_SET = [0.001, 0.002, 0.005, 0.01, 0.02, 0.05, 0.1, 0.2, 0.5, 1, 2] diff --git a/mis_dro/epsilon.py b/mis_dro/epsilon.py new file mode 100644 index 0000000..20ac427 --- /dev/null +++ b/mis_dro/epsilon.py @@ -0,0 +1,56 @@ +import numpy as np +import scipy as sp +from sklearn.model_selection import KFold + +from .bayes_conjugates import default_prior_params, get_posterior_params, get_log_partition_constant, derive_analytical_posterior_params, posterior_predictive_params +from .kl_divergence import kl_divergence_gaussian_kde_monte_carlo + +def get_num_observations_in_train_split(n_splits: int, split_idx: int, n_observations: int): + ratio = float(n_observations) / float(n_splits) + num_splits_with_less_than_max_test_size = n_splits * np.ceil(ratio) - n_observations + if split_idx < n_splits - num_splits_with_less_than_max_test_size: + return int(n_observations - np.ceil(ratio)) + else: + return int(n_observations - np.floor(ratio)) + +def get_kde_epsilon_cross_validation(data: np.array, algorithm: str, posterior: str, likelihood: str, n_splits: int, cv_seed: int) -> float: + cv_random_state = np.random.RandomState(seed=cv_seed) + kf = KFold(n_splits=n_splits, shuffle=True, random_state=cv_random_state) + dim = data.shape[1] + epsilon_values = np.zeros(n_splits) + for i, (train_index, test_index) in enumerate(kf.split(data)): + fold_train = data[train_index] + fold_test = data[test_index] + theta_prior = default_prior_params(posterior, dim=dim) + theta_posterior = get_posterior_params(posterior, fold_train, theta_prior) + if algorithm == "kl_dro_bas": + log_partition_constant = get_log_partition_constant(posterior, theta_posterior) + theta_sample = derive_analytical_posterior_params( + posterior, theta_posterior + ) + model = get_scipy_likelihood_from_theta(theta_sample[0], likelihood) + elif algorithm == "kl_pp": + theta_sample = posterior_predictive_params(posterior, theta_posterior) + model = get_scipy_posterior_predictive(theta_sample[0], posterior, likelihood) + log_partition_constant = 0.0 + else: + raise NotImplementedError() + epsilon_values[i] = kl_divergence_gaussian_kde_monte_carlo(fold_test, model) + log_partition_constant + return epsilon_values.mean() + +def get_scipy_likelihood_from_theta(theta: np.array, likelihood: str): + if likelihood == "normal": + mu, scale = theta + return sp.stats.norm(loc=mu, scale=scale) + if likelihood == "exponential": + return sp.stats.expon(scale=1.0 / theta) + raise NotImplementedError(f"Likelihood: {likelihood}") + +def get_scipy_posterior_predictive(theta: np.array, posterior: str, likelihood: str): + if likelihood == "normal" and posterior == "normal_gamma": + mu, scale, df = theta + return sp.stats.t(df, loc=mu, scale=scale) + if likelihood == "exponential" and posterior == "gamma": + shape, scale = theta + sp.stats.lomax(c=shape, scale=scale) + raise NotImplementedError(f"Posterior predictive for likelihood {likelihood} and posterior {posterior}.") \ No newline at end of file diff --git a/mis_dro/experiments.py b/mis_dro/experiments.py index 2510475..9b9313f 100644 --- a/mis_dro/experiments.py +++ b/mis_dro/experiments.py @@ -26,7 +26,7 @@ IN_SAMPLE_TIME_WINDOW, OUT_OF_SAMPLE_TIME_WINDOW, ROBAS_DRO_EPSILON_SET, - SMALL_BAS_DRO_EPSILON_SET, + WASSERSTEIN_DRO_EPSILON_SET, ) from .dataset import get_num_time_windows, get_portfolio_returns_df @@ -48,11 +48,16 @@ class ExperimentName(StrEnum): kl_portfolio_synthetic = "kl_portfolio_synthetic" kl_newsvendor_exp_1d = "kl_newsvendor_exp_1d" mmd_newsvendor_exp_1d = "mmd_newsvendor_exp_1d" + cv_kl_newsvendor_1d = "cv_kl_newsvendor_1d" + kde_epsilon_newsvendor_1d = "kde_epsilon_newsvendor_1d" + cv_kl_portfolio = "cv_kl_portfolio" + kl_newsvendor_100 = "kl_newsvendor_100" def is_portfolio(self) -> bool: - return self in (ExperimentName.kl_portfolio, ExperimentName.mmd_portfolio, ExperimentName.kl_portfolio_crash, ExperimentName.mmd_portfolio_crash) - + return self in (ExperimentName.kl_portfolio, ExperimentName.mmd_portfolio, ExperimentName.kl_portfolio_crash, ExperimentName.mmd_portfolio_crash, ExperimentName.cv_kl_portfolio) + def is_cross_validation(self) -> bool: + return self in (ExperimentName.cv_kl_newsvendor_1d, ExperimentName.cv_kl_portfolio) def get_experiment(experiment_name: ExperimentName, dataset_dir: Optional[Path] = None) -> List[Dict]: """Returns the experiment associated with the name""" @@ -71,6 +76,10 @@ def get_experiment(experiment_name: ExperimentName, dataset_dir: Optional[Path] ExperimentName.kl_newsvendor_exp_1d: kl_newsvendor_exp_1d, ExperimentName.mmd_newsvendor_exp_1d: mmd_newsvendor_exp_1d, ExperimentName.mmd_portfolio_synthetic: mmd_portfolio_synthetic, + ExperimentName.cv_kl_newsvendor_1d: cv_kl_newsvendor_1d, + ExperimentName.kde_epsilon_newsvendor_1d: kde_epsilon_newsvendor_1d, + ExperimentName.cv_kl_portfolio: cv_kl_portfolio, + ExperimentName.kl_newsvendor_100: kl_newsvendor_100, } try: if experiment_name.is_portfolio(): @@ -132,7 +141,7 @@ def get_num_likelihood_samples(dataset: str, num_observations: int, num_total_sa return int(np.sqrt(num_total_samples)) if algorithm in ("kl_dro_bas", "kl_pp"): return num_total_samples - if algorithm == "kl_empirical": + if algorithm in ("kl_empirical", "wasserstein_empirical"): return num_observations raise NotImplementedError() @@ -143,38 +152,41 @@ def get_num_posterior_samples(dataset: str, num_total_samples: int, algorithm: s return int(np.sqrt(num_total_samples)) if algorithm in ("kl_dro_bas", "kl_pp"): return 1 - if algorithm == "kl_empirical": + if algorithm in ("wasserstein_empirical", "kl_empirical"): return 1 raise NotImplementedError() def kl_newsvendor_1d() -> List[Dict]: """KL univariate newsvendor: compare our Bayesian ambiguity set against Bayesian DRO""" experiment = [] - for algorithm, num_observations, (dgp, likelihood, posterior), epsilon in itertools.product( - ["kl_pp", "kl_dro_bas", "kl_bdro", "kl_empirical"], + for algorithm, num_observations, (dgp, likelihood, posterior) in itertools.product( + ["kl_pp", "kl_dro_bas", "kl_bdro", "kl_empirical", "wasserstein_empirical"], [NUM_OBSERVATIONS], # [5, 20, 100], [ - ("normal", "normal", "normal_gamma"), + # ("normal", "normal", "normal_gamma"), # ("truncated_normal", "normal", "normal_gamma"), ("exponential", "exponential", "gamma"), # ("contaminated_exp", "exponential", "gamma"), ], - BAS_DRO_EPSILON_SET, - # SMALL_BAS_DRO_EPSILON_SET, ): - if algorithm == "kl_empirical": + if algorithm in ("wasserstein_empirical", "kl_empirical"): total_model_samples_list = [0] likelihood = "empirical" posterior = "empirical" inference = "empirical" + if algorithm == "wasserstein_empirical": + epsilon_list = WASSERSTEIN_DRO_EPSILON_SET + elif algorithm == "kl_empirical": + epsilon_list = BAS_DRO_EPSILON_SET else: # total_model_samples_list = BAS_TOTAL_MODEL_SAMPLES total_model_samples_list = [3600, 10000] inference = "bayes" + epsilon_list = BAS_DRO_EPSILON_SET contamination = 0.0 if dgp == "contaminated_exp": contamination = CONTAMINATION_LEVEL - for total_model_samples in total_model_samples_list: + for epsilon, total_model_samples in itertools.product(epsilon_list, total_model_samples_list): params = { "algorithm": algorithm, "contamination": contamination, @@ -190,7 +202,84 @@ def kl_newsvendor_1d() -> List[Dict]: "num_likelihood_samples": get_num_likelihood_samples("newsvendor", num_observations, total_model_samples, algorithm), "num_observations": num_observations, "num_posterior_samples": get_num_posterior_samples("newsvendor", total_model_samples, algorithm), - "num_replications": BAS_NUM_REPLICATIONS, + # "num_replications": BAS_NUM_REPLICATIONS, # FIXME + "num_replications": 200, + "num_test_observations": NUM_TEST_OBSERVATIONS, + "posterior": posterior, + "uuid": str(uuid4()), # uniquely identify a run + } + experiment.append(params) + return experiment + + +def kde_epsilon_newsvendor_1d() -> List[Dict]: + """KL univariate newsvendor: compare our Bayesian ambiguity set against Bayesian DRO""" + experiment = [] + for algorithm, (dgp, likelihood, posterior), total_model_samples in itertools.product( + ["kl_pp", "kl_dro_bas"], + [ + ("normal", "normal", "normal_gamma"), + ("truncated_normal", "normal", "normal_gamma"), + ("exponential", "exponential", "gamma"), + # ("contaminated_exp", "exponential", "gamma"), + ], + BAS_TOTAL_MODEL_SAMPLES, + ): + num_observations = 100 # need more observations to do cross-validation + contamination = 0.0 + if dgp == "contaminated_exp": + contamination = CONTAMINATION_LEVEL + params = { + "algorithm": algorithm, + "contamination": contamination, + "dataset": "newsvendor", + "dgp": dgp, + "dim": 1, + "epsilon": None, # we will use kde epsilon instead! + "kde_epsilon": True, + "n_splits": 10, + "ignore_dpp": True, + "inference": "bayes", + "lengthscale": -1.0, + "likelihood": likelihood, + "njobs": 1, + "num_likelihood_samples": get_num_likelihood_samples("newsvendor", num_observations, total_model_samples, algorithm), + "num_observations": 100, + "num_posterior_samples": get_num_posterior_samples("newsvendor", total_model_samples, algorithm), + "num_replications": BAS_NUM_REPLICATIONS, # FIXME + "num_test_observations": NUM_TEST_OBSERVATIONS, + "posterior": posterior, + "uuid": str(uuid4()), # uniquely identify a run + } + experiment.append(params) + return experiment + +def kl_newsvendor_100() -> List[Dict]: + """KL univariate newsvendor with 100 observations on normal DGP""" + experiment = [] + for algorithm in ["kl_pp", "kl_dro_bas"]: + total_model_samples_list = BAS_TOTAL_MODEL_SAMPLES + dgp, likelihood, posterior = "normal", "normal", "normal_gamma" + inference = "bayes" + num_observations = 100 # we are comparing against CV, which uses more observations + epsilon_list = BAS_DRO_EPSILON_SET + for epsilon, total_model_samples in itertools.product(epsilon_list, total_model_samples_list): + params = { + "algorithm": algorithm, + "dataset": "newsvendor", + "dgp": dgp, + "dim": 1, + "epsilon": epsilon, + "ignore_dpp": True, + "inference": inference, + "lengthscale": -1.0, + "likelihood": likelihood, + "njobs": 1, + "num_likelihood_samples": get_num_likelihood_samples("newsvendor", num_observations, total_model_samples, algorithm), + "num_observations": num_observations, + "num_posterior_samples": get_num_posterior_samples("newsvendor", total_model_samples, algorithm), + # "num_replications": BAS_NUM_REPLICATIONS, # FIXME? + "num_replications": 200, "num_test_observations": NUM_TEST_OBSERVATIONS, "posterior": posterior, "uuid": str(uuid4()), # uniquely identify a run @@ -198,6 +287,133 @@ def kl_newsvendor_1d() -> List[Dict]: experiment.append(params) return experiment + +def cv_kl_newsvendor_1d() -> List[Dict]: + """Cross-validation KL univariate newsvendor for selecting epsilon""" + experiment = [] + for algorithm, num_observations, (dgp, likelihood, posterior) in itertools.product( + ["kl_pp", "kl_dro_bas", "kl_bdro"], + [100], # FIXME? + [ + ("normal", "normal", "normal_gamma"), + ("truncated_normal", "normal", "normal_gamma"), + ("exponential", "exponential", "gamma"), + # ("contaminated_exp", "exponential", "gamma"), + ], + ): + if algorithm in ("wasserstein_empirical", "kl_empirical"): + total_model_samples_list = [0] + likelihood = "empirical" + posterior = "empirical" + inference = "empirical" + if algorithm == "wasserstein_empirical": + epsilon_list = WASSERSTEIN_DRO_EPSILON_SET + elif algorithm == "kl_empirical": + epsilon_list = BAS_DRO_EPSILON_SET + else: + total_model_samples_list = BAS_TOTAL_MODEL_SAMPLES + inference = "bayes" + epsilon_list = BAS_DRO_EPSILON_SET + contamination = 0.0 + if dgp == "contaminated_exp": + contamination = CONTAMINATION_LEVEL + NUM_SPLITS = 10 + for total_model_samples in total_model_samples_list: + base_params = { + "algorithm": algorithm, + "contamination": contamination, + "dataset": "newsvendor", + "dgp": dgp, + "dim": 1, + "ignore_dpp": True, + "inference": inference, + "lengthscale": -1.0, + "likelihood": likelihood, + "njobs": 1, + "num_likelihood_samples": get_num_likelihood_samples("newsvendor", num_observations, total_model_samples, algorithm), + "num_observations": num_observations, + "num_posterior_samples": get_num_posterior_samples("newsvendor", total_model_samples, algorithm), + # "num_replications": BAS_NUM_REPLICATIONS, # FIXME + "num_replications": 200, + "num_test_observations": NUM_TEST_OBSERVATIONS, + "posterior": posterior, + "do_cross_validation": True, + "n_splits": NUM_SPLITS, + } + cv_uuid_list = [] + for epsilon in epsilon_list: + for split_idx in range(NUM_SPLITS): + fold_params = base_params.copy() + fold_params["uuid"] = str(uuid4()) + fold_params["epsilon"] = epsilon + fold_params["n_splits"] = NUM_SPLITS + fold_params["split_idx"] = split_idx + fold_params["use_cv_epsilon"] = False + fold_params["cv_uuid_list"] = [] + experiment.append(fold_params) + cv_uuid_list.append(fold_params["uuid"]) + params = base_params.copy() + params["uuid"] = str(uuid4()) + params["epsilon"] = None # this must be calculated later using CV! + params["split_idx"] = None # not needed because we will calculate epsilon using all splits + params["use_cv_epsilon"] = True # we will exploit the CV epsilon + params["cv_uuid_list"] = cv_uuid_list #NOTE point to all the UUIDs across all folds and epsilons + experiment.append(params) + return experiment + +def cv_kl_portfolio(mmc2_dir: Path) -> List[Dict]: + """Cross-validation KL univariate portfolio for selecting epsilon""" + experiment = [] + dgp = "DowJones" + returns_df = get_portfolio_returns_df(mmc2_dir, dgp) + num_time_windows = get_num_time_windows(len(returns_df)) + num_stocks = len(returns_df.columns) + for algorithm in ["kl_dro_bas", "kl_bdro", "kl_pp"]: + likelihood = "multivariate_normal" + posterior = "normal_inverse_wishart" + inference = "bayes" + total_model_samples = 900 + NUM_SPLITS = 10 + base_params = { + "algorithm": algorithm, + "contamination": 0.0, + "dataset": "portfolio", + "dgp": dgp, + "dim": num_stocks, + "ignore_dpp": True, + "inference": inference, + "likelihood": likelihood, + "njobs": 1, + "num_likelihood_samples": get_num_likelihood_samples("newsvendor", IN_SAMPLE_TIME_WINDOW, total_model_samples, algorithm), + "num_observations": IN_SAMPLE_TIME_WINDOW, + "num_posterior_samples": get_num_posterior_samples("newsvendor", total_model_samples, algorithm), + "num_replications": num_time_windows, + "num_test_observations": OUT_OF_SAMPLE_TIME_WINDOW, + "posterior": posterior, + "do_cross_validation": True, + "n_splits": NUM_SPLITS, + } + cv_uuid_list = [] + for epsilon in PORTFOLIO_EPSILON_SET: + for split_idx in range(NUM_SPLITS): + fold_params = base_params.copy() + fold_params["uuid"] = str(uuid4()) + fold_params["epsilon"] = epsilon + fold_params["n_splits"] = NUM_SPLITS + fold_params["split_idx"] = split_idx + fold_params["use_cv_epsilon"] = False + fold_params["cv_uuid_list"] = [] + experiment.append(fold_params) + cv_uuid_list.append(fold_params["uuid"]) + params = base_params.copy() + params["uuid"] = str(uuid4()) + params["epsilon"] = None # this must be calculated later using CV! + params["split_idx"] = None # not needed because we will calculate epsilon using all splits + params["use_cv_epsilon"] = True # we will exploit the CV epsilon + params["cv_uuid_list"] = cv_uuid_list #NOTE point to all the UUIDs across all folds and epsilons + experiment.append(params) + return experiment + def kl_newsvendor_exp_1d() -> List[Dict]: experiment = [] total_model_samples = 900 diff --git a/mis_dro/kl_divergence.py b/mis_dro/kl_divergence.py index 6550723..a67f45c 100644 --- a/mis_dro/kl_divergence.py +++ b/mis_dro/kl_divergence.py @@ -3,6 +3,8 @@ from typing import Optional import numpy as np import scipy as sp +from sklearn.neighbors import KernelDensity + def kl_divergence_monte_carlo( p: sp.stats.rv_continuous, @@ -31,6 +33,27 @@ def kl_divergence_monte_carlo( # then calculate the expectation using the samples return np.mean(np.log(p.pdf(p_samples) / q.pdf(p_samples))) +def kl_divergence_gaussian_kde_monte_carlo( + observations: np.ndarray, + q: sp.stats.rv_continuous, +) -> float: + """KL divergence approximation using Monte Carlo and Gaussian KDE. + + Args: + observations: Data with shape (num_observations, dim) + q: scipy continuous distribution + + Returns: + KL divergence between observations and q + + Notes: + The KL is the integral of p(x) log(p(x)/q(x)). + This can be written as the expected value under p(x) of the log term. + We sample from p(x) then evaluate the expectation of the log term under these samples. + """ + kde = sp.stats.gaussian_kde(observations.T) + return np.mean(np.log(kde.pdf(observations.T) / q.pdf(observations))) + def kl_divergence_approx_histogram(p_samples, q_samples, nbins=100): all_samples = np.concatenate([p_samples, q_samples]) diff --git a/mis_dro/kl_dro_bas_template.slurm b/mis_dro/kl_dro_bas_template.slurm index 05c245b..09c1cf3 100644 --- a/mis_dro/kl_dro_bas_template.slurm +++ b/mis_dro/kl_dro_bas_template.slurm @@ -14,4 +14,4 @@ ## Loop over each batch ## start=$(($SLURM_ARRAY_TASK_ID * 1)) -srun misdro batch {experiment_dir} $start {batch_size} +srun misdro batch {experiment_dir} $start {batch_size} \ No newline at end of file diff --git a/mis_dro/main.py b/mis_dro/main.py index 70dbe5d..8692f00 100644 --- a/mis_dro/main.py +++ b/mis_dro/main.py @@ -12,6 +12,7 @@ import pandas as pd import scipy as sp import typer +from sklearn.model_selection import KFold from bayesian_dro.Bayesian_DRO_continuous import main_Bayesian_DRO from .bayes_conjugates import ( @@ -37,14 +38,17 @@ ROBAS_NEWSVENDOR_NUM_REPLICATIONS, ) from .dataset import sample_dgp, portfolio_dataset, get_num_time_windows +from .epsilon import get_num_observations_in_train_split, get_kde_epsilon_cross_validation from .experiments import ExperimentName, get_experiment from .likelihood import sample_likelihood, reconstruct_covariance_from_triu -from .newsvendor import newsvendor_cost_cvxpy +from .newsvendor import newsvendor_cost_cvxpy, empirical_wasserstein_dro_newsvendor from .npl import sample_npl from .optimise import get_kl_bdro_problem, DRO_BAS_MMD from .portfolio import get_kl_portfolio_problem, bdro_portfolio_posterior_samples, portfolio_objective_cvxpy from .preprocessing import normalise_by_dimension from .gaussian_kernel import * +from .results import get_result_df_list, preprocess_results_df, get_agg_df, is_minimise_pareto_front, is_maximise_pareto_front, convert_str_to_float_list + app = typer.Typer(name="misdro") @@ -66,17 +70,46 @@ def setup_kl_dro_bas( json.dump(experiment, json_file, indent=4) # for the given batch size, how many batches do we need? - num_batches = math.ceil(float(len(experiment)) / float(batch_size)) + if not experiment_name.is_cross_validation(): + num_batches = math.ceil(float(len(experiment)) / float(batch_size)) - # setup SLURM file - with open( - Path(__file__).parent / "kl_dro_bas_template.slurm", "r", encoding="utf-8" - ) as slurm_file: - slurm_string = slurm_file.read() - dgp_string = slurm_string.format( - experiment_dir=experiment_dir, num_batches=num_batches, batch_size=batch_size - ) - (experiment_dir / f"{experiment_name}.slurm").write_text(dgp_string) + # setup SLURM file + with open( + Path(__file__).parent / "kl_dro_bas_template.slurm", "r", encoding="utf-8" + ) as slurm_file: + slurm_string = slurm_file.read() + dgp_string = slurm_string.format( + experiment_dir=experiment_dir, num_batches=num_batches, batch_size=batch_size + ) + (experiment_dir / f"{experiment_name}.slurm").write_text(dgp_string) + + else: + # separate into two experiments which need separate SLURM files: + # A runs all the splits across all epsilons, + # B uses the epsilons calculated by cross-validation + fold_experiment = [params for params in experiment if params["do_cross_validation"] and not params["use_cv_epsilon"]] + fold_num_batches = math.ceil(float(len(fold_experiment)) / float(batch_size)) + with open( + Path(__file__).parent / "kl_dro_bas_template.slurm", "r", encoding="utf-8" + ) as slurm_file: + slurm_string = slurm_file.read() + fold_string = slurm_string.format( + experiment_dir=experiment_dir, num_batches=fold_num_batches, batch_size=batch_size + ) + fold_string += " --do-cross-validation --no-use-cv-epsilon" + (experiment_dir / f"do_cross_validation.slurm").write_text(fold_string) + + use_cv_epsilon_experiment = [params for params in experiment if params["do_cross_validation"] and params["use_cv_epsilon"]] + use_cv_epsilon_num_batches = math.ceil(float(len(use_cv_epsilon_experiment)) / float(batch_size)) + with open( + Path(__file__).parent / "kl_dro_bas_template.slurm", "r", encoding="utf-8" + ) as slurm_file: + slurm_string = slurm_file.read() + use_cv_epsilon_string = slurm_string.format( + experiment_dir=experiment_dir, num_batches=use_cv_epsilon_num_batches, batch_size=batch_size + ) + use_cv_epsilon_string += " --do-cross-validation --use-cv-epsilon" + (experiment_dir / f"use_cv_epsilon.slurm").write_text(use_cv_epsilon_string) @app.command(name="setup-mmd") @@ -147,25 +180,8 @@ def generate_csv(experiment_dir: Path, npl_samples_dir: Optional[Path] = None): experiment_df = pd.DataFrame(experiment).set_index("uuid") print("Loading and concatenating", len(experiment_df), "CSV files into a pandas dataframe...") result_df = pd.DataFrame() - result_list = [] - failed_uuid_list = [] - missing_uuid_list = [] - for uuid in experiment_df.index: - if (experiment_dir / f"{uuid}.csv").exists(): - try: - result_list.append(pd.read_csv( - experiment_dir / f"{uuid}.csv", index_col=["uuid", "replication"] - )) - except pd.errors.ParserError: - failed_uuid_list.append(uuid) - else: - missing_uuid_list.append(uuid) + result_list = get_result_df_list(experiment_dir, experiment_df.index) - print("The following UUIDs did not have a CSV file:") - print(missing_uuid_list) - print() - print("The following UUIDs failed due to a pandas.errors.ParserError:") - print(failed_uuid_list) result_df = pd.concat([result_df] + result_list) result_df = result_df.join(experiment_df, on="uuid") result_df = result_df.reset_index() @@ -208,12 +224,20 @@ def run_experiment( @app.command(name="batch") -def batch(experiment_dir: Path, batch_id: int, batch_size: int, only_missing: bool = False, dataset_dir: Path = Path("~/datasets/misdro/mmc2"), npl_samples_dir: Optional[Path] = None): +def batch(experiment_dir: Path, batch_id: int, batch_size: int, only_missing: bool = False, dataset_dir: Path = Path("~/datasets/misdro/mmc2"), npl_samples_dir: Optional[Path] = None, do_cross_validation: bool = False, use_cv_epsilon: bool = False): print(datetime.now(), "Running batch from array index", batch_id) print() filepath = experiment_dir / "experiment.json" with open(filepath, "r", encoding="utf-8") as json_file: experiment = json.load(json_file) + if do_cross_validation and not use_cv_epsilon: + # only keep parameters where do_cross_validation is set to True + experiment = [params for params in experiment if params["do_cross_validation"] and not params["use_cv_epsilon"]] + print(len(experiment), "params to run in this cross-validation batch.") + elif do_cross_validation and use_cv_epsilon: + experiment = [params for params in experiment if params["do_cross_validation"] and params["use_cv_epsilon"]] + print(len(experiment), "params to run in this 'use_cv_epsilon' batch.") + start = batch_id * batch_size batch_experiment = experiment[start: min(start + batch_size, len(experiment))] for params in batch_experiment: @@ -245,7 +269,7 @@ def run( dataset_dir: Optional[Path] = None, dgp: str = "truncated_normal", dim: int = 1, - epsilon: float = 1.0, + epsilon: Optional[float] = 1.0, eta: float = NPL_ETA, ignore_dpp: bool = False, inference: str = "bayes", @@ -262,12 +286,107 @@ def run( num_test_observations: int = NUM_TEST_OBSERVATIONS, num_certify_points: int = NUM_CERTIFY, posterior: str = "gamma", + do_cross_validation: bool = False, + n_splits: Optional[int] = None, + split_idx: Optional[int] = None, + use_cv_epsilon: bool = False, + cv_uuid_list: list[str] = [], + kde_epsilon: bool = False, uuid: str = str(uuid4()), verbose: bool = False, ): """Run Newsvendor Misspecified Bayesian DRO""" if uuid: print(uuid) + + if do_cross_validation and n_splits is not None and split_idx is not None and epsilon is not None: + num_training_observations = get_num_observations_in_train_split(n_splits, split_idx, num_observations) + num_test_observations = num_observations - num_training_observations + print(f"Doing {n_splits}-fold cross-validation on split {split_idx}: training/test set size is {num_training_observations}/{num_test_observations}.") + elif use_cv_epsilon and split_idx is None and do_cross_validation: + num_training_observations = num_observations + # TODO get the best epsilon for each replication from the cross-validation - store in array + # load the results df for each UUID in cv_uuid_list + result_list = get_result_df_list(experiment_dir, cv_uuid_list) + result_df = pd.concat(result_list) + result_df["out_of_sample_cost"] = result_df["out_of_sample_cost"].map(lambda x: convert_str_to_float_list(x, num_test_observations)) + + experiment_filepath = experiment_dir / "experiment.json" + with open(experiment_filepath, "r", encoding="utf-8") as json_file: + experiment = json.load(json_file) + experiment_df = pd.DataFrame(experiment).set_index("uuid") + result_df = result_df.join(experiment_df, on="uuid") + + epsilons_for_replications = np.zeros(num_replications) + + # each replication may have a different epsilon + # for replication in range(num_replications): + # replication_df = result_df.loc[result_df.index.get_level_values("replication") == replication] + # best_epsilon_for_each_split = np.zeros(n_splits) + # # for each split, get the epsilon that achieves the minimum OOS cost + # for split in range(n_splits): + # split_df = replication_df.loc[replication_df["split_idx"] == split] + # split_df["out_of_sample_mean"] = split_df["out_of_sample_cost"].apply(np.mean).values + # if dataset == "newsvendor": + # epsilon = split_df.loc[split_df["out_of_sample_mean"]==split_df["out_of_sample_mean"].min()].iloc[0]["epsilon"] + # elif dataset == "portfolio": + # epsilon = split_df.loc[split_df["out_of_sample_mean"]==split_df["out_of_sample_mean"].max()].iloc[0]["epsilon"] + # best_epsilon_for_each_split[split] = epsilon + # print("split =",split, ". Epsilon =", epsilon) + # # then take the average of the epsilon values that achieve this minimum + # epsilons_for_replications[replication] = np.median(best_epsilon_for_each_split) + # print("Best epsilon for replication is", epsilons_for_replications[replication]) + + # group by replication and epsilon and get the OOS mean and variance + # gb = result_df.groupby(["epsilon", "replication"]) + # agg_df = gb.agg( + # out_of_sample_mean = pd.NamedAgg(column="out_of_sample_cost", aggfunc=lambda x: np.mean(np.concatenate(x.values))), + # out_of_sample_var = pd.NamedAgg(column="out_of_sample_cost", aggfunc=lambda x: np.var(np.concatenate(x.values), ddof=1)), + # ) + # for replication in range(num_replications): + # replication_df = agg_df.loc[agg_df.index.get_level_values("replication") == replication] + # if dataset == "newsvendor": + # epsilons_for_replications[replication] = replication_df.loc[replication_df["out_of_sample_mean"] == replication_df["out_of_sample_mean"].min()].index[0][0] + # elif dataset == "portfolio": + # epsilons_for_replications[replication] = replication_df.loc[replication_df["out_of_sample_mean"] == replication_df["out_of_sample_mean"].max()].index[0][0] + + assert len(result_df["algorithm"].unique()) == 1 + gb = result_df.groupby(["epsilon"]) + agg_df = gb.agg( + out_of_sample_mean = pd.NamedAgg(column="out_of_sample_cost", aggfunc=lambda x: np.mean(np.concatenate(x.values))), + out_of_sample_var = pd.NamedAgg(column="out_of_sample_cost", aggfunc=lambda x: np.var(np.concatenate(x.values), ddof=1)), + ) + + if dataset == "newsvendor": + epsilon = agg_df.loc[agg_df["out_of_sample_mean"] == agg_df["out_of_sample_mean"].min()].index[0] + elif dataset == "portfolio": + epsilon = agg_df.loc[agg_df["out_of_sample_mean"] == agg_df["out_of_sample_mean"].max()].index[0] + print("Epsilon:", epsilon) + + + # for replication in range(num_replications): + # replication_df = agg_df.loc[agg_df.index.get_level_values("replication") == replication] + # assert len(replication_df) + # if dataset == "newsvendor": + # is_pareto_front = is_minimise_pareto_front(replication_df["out_of_sample_var"].values, replication_df["out_of_sample_mean"].values) + # # one could do multiple things here - e.g. choose smallest OOS mean, smallest OOS variance, etc. + # elif dataset == "portfolio": + # is_pareto_front = is_maximise_pareto_front(replication_df["out_of_sample_var"].values, replication_df["out_of_sample_mean"].values) + # else: + # raise NotImplementedError(dataset) + # assert len(is_pareto_front), "There is not at least one pareto optimal point" + # pareto_df = replication_df[is_pareto_front] + # # take the mean of the epsilons in the Pareto df + # print("Pareto frontier:") + # print(pareto_df) + # epsilons_for_replications[replication] = np.mean(pareto_df.index.get_level_values("epsilon")) + + + elif do_cross_validation: + raise ValueError("Something went wrong in the previous logic.") + else: + num_training_observations = num_observations + print("DGP:", dgp, " - ALGORITHM:", algorithm, " - NUM LIKELIHOOD SAMPLES:", num_likelihood_samples, " - POSTERIOR:", posterior, "- DATASET:", dataset, "- DIM:", dim) if algorithm in ("kl_bdro", "kl_dro_bas", "kl_pp", "kl_empirical") and dataset == "newsvendor": problem = get_kl_bdro_problem( @@ -276,7 +395,7 @@ def run( elif algorithm == "kl_pp" and dataset in ("portfolio", "portfolio_synthetic"): problem = get_kl_bdro_problem(portfolio_objective_cvxpy, num_posterior_samples, num_likelihood_samples, dim=dim, is_portfolio=True) elif algorithm == "kl_empirical" and dataset in ("portfolio", "portfolio_synthetic"): - problem = get_kl_bdro_problem(portfolio_objective_cvxpy, 1, num_observations, dim=dim, is_portfolio=True) + problem = get_kl_bdro_problem(portfolio_objective_cvxpy, 1, num_training_observations, dim=dim, is_portfolio=True) elif algorithm in ("kl_bdro", "kl_dro_bas") and dataset in ("portfolio", "portfolio_synthetic") and likelihood == "multivariate_normal": problem = get_kl_portfolio_problem(dim, num_posterior_samples) elif algorithm in ("dro_bas_mmd", "empirical_mmd"): @@ -284,7 +403,7 @@ def run( if algorithm == "dro_bas_mmd": n_samples = num_posterior_samples*num_likelihood_samples elif algorithm == "empirical_mmd": - n_samples = num_observations + n_samples = num_training_observations if dataset == "newsvendor": kdro_class = DRO_BAS_MMD(dim_theta, dim, newsvendor_cost_cvxpy) problem = kdro_class.get_newsvendor_problem(n_samples, num_certify_points) @@ -293,6 +412,9 @@ def run( problem = kdro_class.get_portfolio_problem(n_samples, num_certify_points) else: raise ValueError(f"Objective not implemented for dataset '{dataset}'") + elif algorithm == "wasserstein_empirical": + # NOTE we don't need a problem here - solution is found using bisection search + problem = None else: raise NotImplementedError(f"Algorithm {algorithm} not implemented.") # If the number of parameters is small enough, then use Disciplined Parametrized Programming (DPP) @@ -318,6 +440,7 @@ def run( "eta": eta, "ignore_dpp": ignore_dpp, "inference": inference, + "kernel_name": kernel_name, "lengthscale": lengthscale, "likelihood": likelihood, "normalise": normalise, @@ -328,6 +451,11 @@ def run( "num_replications": num_replications, "num_test_observations": num_test_observations, "posterior": posterior, + "do_cross_validation": do_cross_validation, + "n_splits": n_splits, + "split_idx": split_idx, + "use_cv_epsilon": use_cv_epsilon, + "kde_epsilon": kde_epsilon, "uuid": uuid, "verbose": verbose, } @@ -340,12 +468,19 @@ def run( else: params["npl_uuid_dir"] = None params.pop("num_replications") # popped because we don't need to pass this to the run_replication method, but it is needed above for getting the npl_uuid + # we don't need the following parameters to run optimisation - they are only used for sampling from NPL + params.pop("lengthscale", None) + params.pop("kernel_name", None) + params.pop("eta", None) if njobs == 1: all_solve_start = datetime.now() list_of_replication_stats = [] print(all_solve_start, "- Running all replications in series.") for j in range(num_replications): + if use_cv_epsilon: + # params["epsilon"] = epsilons_for_replications[j] + params["epsilon"] = epsilon list_of_replication_stats.append(run_replication(j, problem, **params)) all_solve_end = datetime.now() print(all_solve_end, "- Finished solving all replications in series. Total solve time is", (all_solve_end - all_solve_start).total_seconds()) @@ -383,11 +518,9 @@ def run_replication( dataset_dir: Optional[Path] = None, dgp: str = "truncated_normal", dim: int = 1, - epsilon: float = 1.0, - eta: float = NPL_ETA, + epsilon: Optional[float] = 1.0, # pass None if using kde_epsilon ignore_dpp: bool = False, inference: str = "bayes", - lengthscale: float = -1.0, likelihood: str = "exponential", normalise: bool = False, npl_uuid_dir: Optional[Path] = None, @@ -397,6 +530,11 @@ def run_replication( num_posterior_samples: int = NUM_POSTERIOR_SAMPLES, num_test_observations: int = NUM_TEST_OBSERVATIONS, posterior: str = "gamma", + do_cross_validation: bool = False, + n_splits: Optional[int] = None, + split_idx: Optional[int] = None, + use_cv_epsilon: bool = False, + kde_epsilon: bool = False, uuid: str = str(uuid4()), verbose: bool = False, ): @@ -425,8 +563,26 @@ def run_replication( data = normalise_by_dimension(data) else: raise NotImplementedError(f"Dataset not implemented: {dataset}") + + if do_cross_validation and not use_cv_epsilon: + # NOTE we use a different random number generator for CV because we do not want to contaminate the test samples + # and because we want to reproduce the same CV splits for each replication + cv_random_state = np.random.RandomState(seed=replication + 1000) + kf = KFold(n_splits=n_splits, shuffle=True, random_state=cv_random_state) + train_index, test_index = list(kf.split(data))[split_idx] + data_eval = data[test_index] + data = data[train_index] + dgp_time = (datetime.now() - dgp_start).total_seconds() + # try to find the best epsilon using a KDE estimate of the empirical distribution + # and the Monte-Carlo approximation of the KL divergence + if kde_epsilon and inference == "bayes" and algorithm in ("kl_pp", "kl_dro_bas"): + assert n_splits is not None + cv_seed = 2000 + replication + epsilon = get_kde_epsilon_cross_validation(data, algorithm, posterior, likelihood, n_splits, cv_seed) + + # 2. sample from the posterior posterior_start = datetime.now() log_partition_constant = 0.0 @@ -481,6 +637,18 @@ def run_replication( ) likelihood_time = (datetime.now() - likelihood_start).total_seconds() + if algorithm == "kl_dro_bas" and ( + dataset == "portfolio" or dataset == "portfolio_synthetic" + or (dataset == "newsvendor" and do_cross_validation) + ): + # NOTE under the above conditions, having values of epsilon just above + # the constant is benefitial for obtaining a small mean + epsilon_prime = epsilon + else: + # as in Corollary 3.7 + # NOTE for BDRO and BAS-PP this is just equal to epsilon because log_partition_constant is zero + epsilon_prime = epsilon - log_partition_constant + # 4. run the chosen DRO algorithm solve_start = datetime.now() solution = np.nan @@ -489,13 +657,7 @@ def run_replication( and algorithm in ("kl_bdro", "kl_dro_bas") and likelihood == "multivariate_normal" ): - # if epsilon - log_partition_constant < 0: - # # NOTE the optimisation problem is unbounded below - # solution = np.inf * np.ones(dim) - # solve_time = 0.0 - # setup_time = 0.0 - # else: - problem.param_dict["epsilon_minus_constant"].value = np.array([epsilon]) + problem.param_dict["epsilon_minus_constant"].value = np.array([epsilon_prime]) problem.param_dict["mu_post"].value = theta_sample[0, :dim] for i in range(num_posterior_samples): # get a PSD covariance from the upper triangular vector @@ -515,7 +677,7 @@ def run_replication( setup_time = 0.0 else: # set parameters then solve - problem.param_dict["epsilon_minus_constant"].value = np.array([epsilon - log_partition_constant]) + problem.param_dict["epsilon_minus_constant"].value = np.array([epsilon_prime]) xi = xi.reshape((num_posterior_samples, num_likelihood_samples, dim)) for i in range(num_posterior_samples): problem.param_dict[f"xi_{i}"].value = xi[i] @@ -548,6 +710,9 @@ def run_replication( elif algorithm == "bdro_grid_search": solution = main_Bayesian_DRO(xi, epsilon) setup_time = 0.0 # can't really measure this easily + elif algorithm == "wasserstein_empirical": + setup_time = 0.0 + solution = np.array([empirical_wasserstein_dro_newsvendor(xi, epsilon, p=2)]) else: raise ValueError("Please choose a valid algorithm") solve_time = (datetime.now() - solve_start).total_seconds() diff --git a/mis_dro/newsvendor.py b/mis_dro/newsvendor.py index a6137ec..4c861f7 100644 --- a/mis_dro/newsvendor.py +++ b/mis_dro/newsvendor.py @@ -25,4 +25,33 @@ def newsvendor_cost_cvxpy(x, xi): X = cp.vstack([x for _ in range(xi.shape[0])]) return cp.maximum(0, X - xi) @ h + cp.maximum(0, xi - X) @ b - +def empirical_wasserstein_dro_newsvendor(data, epsilon, p: int = 2, b: float = BACKORDER_COST, h: float = HOLDING_COST): + """Univariate empirical Wasserstein distributionally robust newsvendor problem""" + if len(data.shape) > 1 and data.shape[1] > 1: + raise ValueError("Data must be univariate") + if len(data.shape) == 2: + data = data.flatten() + len_data = len(data) + if p == 1: + # put data in ascending order + raise NotImplementedError("I don't trust this yet - how is epsilon used?") + ascend_data = np.sort(data) + for i in range(1, len_data + 1): + if (i - 1) / len_data < b / (h + b) and i / len_data >= b / (h + b): + return ascend_data[i - 1] + if p > 1: + Delta = ( + 1 + / (h + b) + * (1 / p) ** (1 / (p - 1)) + * ((p - 1) / p) + * (b ** (p / (p - 1)) - h ** (p / (p - 1))) + ) + Lambda = (1 / (h + b)) * (b ** (p / (p - 1)) * h + h ** (p / (p - 1)) * b) + ascend_data = np.sort(data) + # NOTE appears to do a bisection search here - wonder where this comes from? + for i in range(1, len_data + 1): + if (i - 1) / len_data < b / (h + b) and i / len_data >= b / (h + b): + temp = ascend_data[i - 1] + break + return temp + Delta * p ** (1 / (p - 1)) * epsilon * (1 / Lambda) ** (1 / p) diff --git a/mis_dro/plot.py b/mis_dro/plot.py index cb1b8d0..29fcf95 100644 --- a/mis_dro/plot.py +++ b/mis_dro/plot.py @@ -17,6 +17,7 @@ class AlgorithmLineStyle(StrEnum): dro_bas_mmd = "dotted" empirical_mmd = "dotted" kl_pp = "dotted" + wasserstein_empirical = "dotted" # Color blind palette from https://gist.github.com/thriveth/8560036 CB_color_cycle = [ @@ -38,6 +39,7 @@ class AlgorithmColor(StrEnum): kl_bdro = "#ff7f00" kl_dro_bas = "#984ea3" kl_empirical = "#999999" + wasserstein_empirical = "#a65628" empirical_mmd = "#999999" kl_pp = "#4daf4a" @@ -53,6 +55,7 @@ class AlgorithmName(StrEnum): kl_dro_bas = "DRO-BAS$_{PE}$" kl_pp = "DRO-BAS$_{PP}$" kl_empirical = "Empirical KL" + wasserstein_empirical = "Empirical Wasserstein" @@ -97,6 +100,7 @@ class InferencePrettyName(StrEnum): ("dro_bas_mmd", "npl_mmd"): "*", ("empirical_mmd", "empirical"): "+", ("kl_empirical", "empirical"): "x", + ("wasserstein_empirical", "empirical"): "+", } @@ -124,6 +128,7 @@ def mean_variance_plot( var_col: float = "out_of_sample_var", add_log_partition_function: bool = False, minimise: bool = True, + add_end_epsilons: bool = True, **kwargs, ) -> None: """Plot mean-variance trade-off.""" @@ -148,8 +153,10 @@ def mean_variance_plot( # label the points with epsilon values if is_labelled: + if add_end_epsilons: + special_epsilons += [epsilon_list[0], epsilon_list[-1]] for i, epsilon in enumerate(epsilon_list): - if i == 0 or i == len(epsilon_list) - 1 or epsilon in special_epsilons: + if epsilon in special_epsilons: # if line is blue then put text on bottom left if kwargs["color"] in (AlgorithmColor.kl_dro_bas, AlgorithmColor.kl_empirical): ha = "right" @@ -177,83 +184,4 @@ def mean_variance_plot( ha=ha, va=va, color=kwargs["color"], - ) - -def is_minimise_pareto_front(out_of_sample_var, out_of_sample_mean): - """Returns true if the point lies on the Pareto front of a minimisation problem""" - assert out_of_sample_var.shape == out_of_sample_mean.shape - pareto = [] - for i in range(out_of_sample_var.shape[0]): - point_is_pareto = True - for j in range(out_of_sample_var.shape[0]): - if out_of_sample_var[j] < out_of_sample_var[i] and out_of_sample_mean[j] < out_of_sample_mean[i]: - point_is_pareto = False - break - pareto.append(point_is_pareto) - return pareto - -def is_maximise_pareto_front(out_of_sample_var, out_of_sample_mean): - """Returns true if the point lies on the Pareto front of a maximisation problem""" - assert out_of_sample_var.shape == out_of_sample_mean.shape - pareto = [] - for i in range(out_of_sample_var.shape[0]): - point_is_pareto = True - for j in range(out_of_sample_var.shape[0]): - if out_of_sample_var[j] < out_of_sample_var[i] and out_of_sample_mean[j] > out_of_sample_mean[i]: - point_is_pareto = False - break - pareto.append(point_is_pareto) - return pareto - -def get_agg_df(results_df: pd.DataFrame, gb_cols: list[str]): - """Groupby the given columns then apply summary statistics for each group""" - assert len(results_df["num_replications"].unique()) == 1 - assert len(results_df["num_test_observations"].unique()) == 1 - num_replications = results_df["num_replications"].unique()[0] - num_test_observations = results_df["num_test_observations"].unique()[0] - gb = results_df.groupby(by=gb_cols) - agg_df = gb.agg( - out_of_sample_mean = pd.NamedAgg(column="out_of_sample_cost", aggfunc=lambda x: np.mean(np.concatenate(x.values))), - out_of_sample_var = pd.NamedAgg(column="out_of_sample_cost", aggfunc=lambda x: np.var(np.concatenate(x.values), ddof=1)), - sum_of_in_group_var = pd.NamedAgg(column="in_group_var", aggfunc=lambda x: float(num_test_observations - 1) / float(num_replications * num_test_observations - 1) * np.sum(x.values)), - var_of_in_group_mean = pd.NamedAgg(column="in_group_mean", aggfunc=lambda x: float(num_test_observations * (num_replications - 1)) / float(num_replications * num_test_observations - 1) * np.var(x, ddof=1)), - mean_solve_time = pd.NamedAgg(column="solve_time", aggfunc=np.mean), - std_solve_time = pd.NamedAgg(column="solve_time", aggfunc=np.std), - mean_sample_time = pd.NamedAgg(column="sample_time", aggfunc=np.mean), - std_sample_time = pd.NamedAgg(column="sample_time", aggfunc=np.std), - ) - return agg_df - -def convert_str_to_float_list(str_list: str, list_len: int) -> list[float]: - if str_list == "[]": - return [np.nan for _ in range(list_len)] - else: - return [float(x) for x in str_list.strip('[]').split(',')] - -def preprocess_results_df(results_df: pd.DataFrame, dgp: str, dataset: str = "newsvendor"): - """Filter results, process columns, and create new columns""" - assert len(results_df["num_test_observations"].unique()) == 1 - assert len(results_df.loc[(results_df["dgp"] == dgp)]["dim"].unique()) == 1 - num_test_observations = results_df["num_test_observations"].unique()[0] - - processed_df = results_df.copy() - dim = processed_df.loc[(processed_df["dgp"] == dgp)]["dim"].unique() - # filter by the DGP and cases where the the log partition function is feasible for epsilon - processed_df = processed_df.loc[processed_df["dgp"] == dgp] - if dataset != "portfolio": - processed_df = processed_df.loc[processed_df["log_partition_constant"] < processed_df["epsilon"]] - - # get useful stats such as the number of samples and total time spent sampling - processed_df["num_total_samples"] = processed_df["num_posterior_samples"] * processed_df["num_likelihood_samples"] - processed_df["sample_time"] = processed_df["likelihood_time"] + processed_df["posterior_time"] - - # convert strings into list of floats where necessary - processed_df["out_of_sample_cost"] = processed_df["out_of_sample_cost"].map(lambda x: convert_str_to_float_list(x, num_test_observations)) - processed_df["solution"] = processed_df["solution"].map(lambda x: convert_str_to_float_list(x, dim)) - - # calculate the in-group mean and in-group variance for each replication - processed_df["in_group_mean"] = processed_df["out_of_sample_cost"].map(np.mean) - processed_df["in_group_var"] = processed_df["out_of_sample_cost"].map(lambda x: np.var(x, ddof=1)) - - # return preprocessed dataframe - return processed_df + ) \ No newline at end of file diff --git a/mis_dro/results.py b/mis_dro/results.py new file mode 100644 index 0000000..b418225 --- /dev/null +++ b/mis_dro/results.py @@ -0,0 +1,110 @@ + +from pathlib import Path +import numpy as np +import pandas as pd + +def get_result_df_list(experiment_dir: Path, uuid_list: list[str]): + result_list = [] + failed_uuid_list = [] + missing_uuid_list = [] + for uuid in uuid_list: + if (experiment_dir / f"{uuid}.csv").exists(): + try: + result_list.append(pd.read_csv( + experiment_dir / f"{uuid}.csv", index_col=["uuid", "replication"] + )) + except pd.errors.ParserError: + failed_uuid_list.append(uuid) + else: + missing_uuid_list.append(uuid) + + print("The following UUIDs did not have a CSV file:") + print(missing_uuid_list) + print() + print("The following UUIDs failed due to a pandas.errors.ParserError:") + print(failed_uuid_list) + return result_list + +def preprocess_results_df(results_df: pd.DataFrame, dgp: str, dataset: str = "newsvendor"): + """Filter results, process columns, and create new columns""" + assert len(results_df) + assert len(results_df["num_test_observations"].unique()) == 1 + assert len(results_df.loc[(results_df["dgp"] == dgp)]["dim"].unique()) == 1 + num_test_observations = results_df["num_test_observations"].unique()[0] + + processed_df = results_df.copy() + dim = processed_df.loc[(processed_df["dgp"] == dgp)]["dim"].unique() + # filter by the DGP and cases where the the log partition function is feasible for epsilon + processed_df = processed_df.loc[processed_df["dgp"] == dgp] + if dataset != "portfolio": + if "use_cv_epsilon" not in processed_df.columns: + processed_df["use_cv_epsilon"] = False + processed_df = processed_df.loc[(processed_df["log_partition_constant"] < processed_df["epsilon"]) | (processed_df["use_cv_epsilon"])] + + # get useful stats such as the number of samples and total time spent sampling + processed_df["num_total_samples"] = processed_df["num_posterior_samples"] * processed_df["num_likelihood_samples"] + processed_df["sample_time"] = processed_df["likelihood_time"] + processed_df["posterior_time"] + + # convert strings into list of floats where necessary + processed_df["out_of_sample_cost"] = processed_df["out_of_sample_cost"].map(lambda x: convert_str_to_float_list(x, num_test_observations)) + processed_df["solution"] = processed_df["solution"].map(lambda x: convert_str_to_float_list(x, dim)) + + # calculate the in-group mean and in-group variance for each replication + processed_df["in_group_mean"] = processed_df["out_of_sample_cost"].map(np.mean) + processed_df["in_group_var"] = processed_df["out_of_sample_cost"].map(lambda x: np.var(x, ddof=1)) + + # return preprocessed dataframe + return processed_df + +def get_agg_df(results_df: pd.DataFrame, gb_cols: list[str]): + """Groupby the given columns then apply summary statistics for each group""" + assert len(results_df["num_replications"].unique()) == 1 + assert len(results_df["num_test_observations"].unique()) == 1 + num_replications = results_df["num_replications"].unique()[0] + num_test_observations = results_df["num_test_observations"].unique()[0] + gb = results_df.groupby(by=gb_cols) + agg_df = gb.agg( + out_of_sample_mean = pd.NamedAgg(column="out_of_sample_cost", aggfunc=lambda x: np.mean(np.concatenate(x.values))), + out_of_sample_var = pd.NamedAgg(column="out_of_sample_cost", aggfunc=lambda x: np.var(np.concatenate(x.values), ddof=1)), + out_of_sample_std = pd.NamedAgg(column="out_of_sample_cost", aggfunc=lambda x: np.std(np.concatenate(x.values), ddof=1)), + sum_of_in_group_var = pd.NamedAgg(column="in_group_var", aggfunc=lambda x: float(num_test_observations - 1) / float(num_replications * num_test_observations - 1) * np.sum(x.values)), + var_of_in_group_mean = pd.NamedAgg(column="in_group_mean", aggfunc=lambda x: float(num_test_observations * (num_replications - 1)) / float(num_replications * num_test_observations - 1) * np.var(x, ddof=1)), + mean_solve_time = pd.NamedAgg(column="solve_time", aggfunc=np.mean), + std_solve_time = pd.NamedAgg(column="solve_time", aggfunc=np.std), + mean_sample_time = pd.NamedAgg(column="sample_time", aggfunc=np.mean), + std_sample_time = pd.NamedAgg(column="sample_time", aggfunc=np.std), + ) + return agg_df + +def convert_str_to_float_list(str_list: str, list_len: int) -> list[float]: + if str_list == "[]": + return [np.nan for _ in range(list_len)] + else: + return [float(x) for x in str_list.strip('[]').split(',')] + + +def is_minimise_pareto_front(out_of_sample_var, out_of_sample_mean): + """Returns true if the point lies on the Pareto front of a minimisation problem""" + assert out_of_sample_var.shape == out_of_sample_mean.shape + pareto = [] + for i in range(out_of_sample_var.shape[0]): + point_is_pareto = True + for j in range(out_of_sample_var.shape[0]): + if out_of_sample_var[j] < out_of_sample_var[i] and out_of_sample_mean[j] < out_of_sample_mean[i]: + point_is_pareto = False + break + pareto.append(point_is_pareto) + return pareto + +def is_maximise_pareto_front(out_of_sample_var, out_of_sample_mean): + """Returns true if the point lies on the Pareto front of a maximisation problem""" + assert out_of_sample_var.shape == out_of_sample_mean.shape + pareto = [] + for i in range(out_of_sample_var.shape[0]): + point_is_pareto = True + for j in range(out_of_sample_var.shape[0]): + if out_of_sample_var[j] < out_of_sample_var[i] and out_of_sample_mean[j] > out_of_sample_mean[i]: + point_is_pareto = False + break + pareto.append(point_is_pareto) + return pareto diff --git a/notebooks/cross_validation.ipynb b/notebooks/cross_validation.ipynb new file mode 100644 index 0000000..03dc810 --- /dev/null +++ b/notebooks/cross_validation.ipynb @@ -0,0 +1,96 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": 1, + "metadata": {}, + "outputs": [], + "source": [ + "from sklearn.model_selection import KFold\n", + "import numpy as np" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "n_samples = 20\n", + "dim = 10\n", + "\n", + "def get_num_observations_in_train_split(kf: KFold, split_idx: int, n_observations: int):\n", + " ratio = float(n_observations) / float(kf.n_splits)\n", + " num_splits_with_less_than_max_test_size = kf.n_splits * np.ceil(ratio) - n_observations\n", + " if split_idx < kf.n_splits - num_splits_with_less_than_max_test_size:\n", + " return n_observations - np.ceil(ratio)\n", + " else:\n", + " return n_observations - np.floor(ratio)\n", + "\n", + "for j in range(100):\n", + " generator = np.random.RandomState(j)\n", + " data = np.random.rand(n_samples, dim)\n", + " kf = KFold(n_splits=5, shuffle=True, random_state=generator)\n", + " ratio = float(n_samples) / float(kf.n_splits)\n", + " first_split_test_size = np.ceil(ratio)\n", + " for i, (train_index, test_index) in enumerate(kf.split(data)):\n", + " num_splits_with_less_than_max_test_size = kf.n_splits * first_split_test_size - n_samples\n", + " training_set_size = get_num_observations_in_train_split(kf, i, n_samples)\n", + " assert training_set_size == len(train_index)\n", + " assert len(test_index) == n_samples - training_set_size\n" + ] + }, + { + "cell_type": "code", + "execution_count": 40, + "metadata": {}, + "outputs": [ + { + "data": { + "text/plain": [ + "(array([ 0, 1, 2, 3, 4, 5, 6, 8, 9, 10, 11, 12, 14, 15, 16, 17]),\n", + " array([ 7, 13, 18, 19]))" + ] + }, + "execution_count": 40, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "generator = np.random.RandomState(j)\n", + "data = np.random.rand(n_samples, dim)\n", + "kf = KFold(n_splits=5, shuffle=True, random_state=generator)\n", + "list(kf.split(data))[2]" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [] + } + ], + "metadata": { + "kernelspec": { + "display_name": "mis-dro", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.11.7" + } + }, + "nbformat": 4, + "nbformat_minor": 2 +} diff --git a/notebooks/cv_newsvendor.ipynb b/notebooks/cv_newsvendor.ipynb new file mode 100644 index 0000000..36b953c --- /dev/null +++ b/notebooks/cv_newsvendor.ipynb @@ -0,0 +1,1096 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": 10, + "metadata": {}, + "outputs": [], + "source": [ + "import json\n", + "from pathlib import Path\n", + "import numpy as np\n", + "import pandas as pd\n", + "import scipy as sp\n", + "import matplotlib.pyplot as plt\n", + "\n", + "from mis_dro.plot import *\n", + "from mis_dro.experiments import ExperimentName\n", + "from mis_dro.results import preprocess_results_df, is_minimise_pareto_front, get_agg_df, get_result_df_list, convert_str_to_float_list\n", + "\n", + "\n", + "from matplotlib import rc\n", + "rc('font', **{'family': 'serif', 'serif': ['Computer Modern'], \"size\":14})\n", + "rc('text', usetex=True)" + ] + }, + { + "cell_type": "code", + "execution_count": 11, + "metadata": {}, + "outputs": [], + "source": [ + "# cv_kl_newsvendor_dir = Path(\"/dcs/large/u1508153/misdro/cv_newsvendor_different_replications\")\n", + "# cv_kl_newsvendor_dir = Path(\"/Users/patrick/Experiments/misdro/2025_03_28_cross_validation/cv_newsvendor_different_replications\")\n", + "cv_kl_newsvendor_dir = Path(\"/Users/patrick/Experiments/misdro/2025_03_28_cross_validation/cv_kl_newsvendor_1d\")\n", + "\n", + "cv_kl_newsvendor_df = pd.read_csv(cv_kl_newsvendor_dir / \"results.csv\", index_col=[\"uuid\", \"replication\"])\n", + "\n", + "cv_kl_newsvendor_df = cv_kl_newsvendor_df.loc[cv_kl_newsvendor_df[\"algorithm\"].isin(['kl_pp', 'kl_dro_bas'])]\n", + "\n", + "splits_df = cv_kl_newsvendor_df.loc[~cv_kl_newsvendor_df[\"use_cv_epsilon\"]]\n", + "cv_df = cv_kl_newsvendor_df.loc[cv_kl_newsvendor_df[\"use_cv_epsilon\"]]" + ] + }, + { + "cell_type": "code", + "execution_count": 12, + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
solutiondgp_timelikelihood_timeposterior_timesolve_timesetup_timelog_partition_constantout_of_sample_costalgorithmcontamination...do_cross_validationn_splitsepsilonsplit_idxuse_cv_epsiloncv_uuid_listnum_total_samplessample_timein_group_meanin_group_var
uuidreplication
ff656d45-d92f-4977-8077-7bcff1c296790[31.452356647780974]0.0032740.0001370.0001070.140313NaN0.0[35.42715113817625, 67.85563843541308, 105.201...kl_pp0.0...True100.0010.0False[]250.00024456.0783231044.309123
1[26.45344288553397]0.0007020.0001050.0000670.030750NaN0.0[43.45504560473274, 34.86190525143647, 4.11606...kl_pp0.0...True100.0010.0False[]250.00017234.224435314.706175
2[30.25922983644443]0.0006790.0001070.0000650.030499NaN0.0[5.860559265273505, 25.242872784764558, 63.873...kl_pp0.0...True100.0010.0False[]250.00017243.0699291661.110909
\n", + "

3 rows × 34 columns

\n", + "
" + ], + "text/plain": [ + " solution \\\n", + "uuid replication \n", + "ff656d45-d92f-4977-8077-7bcff1c29679 0 [31.452356647780974] \n", + " 1 [26.45344288553397] \n", + " 2 [30.25922983644443] \n", + "\n", + " dgp_time likelihood_time \\\n", + "uuid replication \n", + "ff656d45-d92f-4977-8077-7bcff1c29679 0 0.003274 0.000137 \n", + " 1 0.000702 0.000105 \n", + " 2 0.000679 0.000107 \n", + "\n", + " posterior_time solve_time \\\n", + "uuid replication \n", + "ff656d45-d92f-4977-8077-7bcff1c29679 0 0.000107 0.140313 \n", + " 1 0.000067 0.030750 \n", + " 2 0.000065 0.030499 \n", + "\n", + " setup_time \\\n", + "uuid replication \n", + "ff656d45-d92f-4977-8077-7bcff1c29679 0 NaN \n", + " 1 NaN \n", + " 2 NaN \n", + "\n", + " log_partition_constant \\\n", + "uuid replication \n", + "ff656d45-d92f-4977-8077-7bcff1c29679 0 0.0 \n", + " 1 0.0 \n", + " 2 0.0 \n", + "\n", + " out_of_sample_cost \\\n", + "uuid replication \n", + "ff656d45-d92f-4977-8077-7bcff1c29679 0 [35.42715113817625, 67.85563843541308, 105.201... \n", + " 1 [43.45504560473274, 34.86190525143647, 4.11606... \n", + " 2 [5.860559265273505, 25.242872784764558, 63.873... \n", + "\n", + " algorithm contamination \\\n", + "uuid replication \n", + "ff656d45-d92f-4977-8077-7bcff1c29679 0 kl_pp 0.0 \n", + " 1 kl_pp 0.0 \n", + " 2 kl_pp 0.0 \n", + "\n", + " ... do_cross_validation \\\n", + "uuid replication ... \n", + "ff656d45-d92f-4977-8077-7bcff1c29679 0 ... True \n", + " 1 ... True \n", + " 2 ... True \n", + "\n", + " n_splits epsilon split_idx \\\n", + "uuid replication \n", + "ff656d45-d92f-4977-8077-7bcff1c29679 0 10 0.001 0.0 \n", + " 1 10 0.001 0.0 \n", + " 2 10 0.001 0.0 \n", + "\n", + " use_cv_epsilon cv_uuid_list \\\n", + "uuid replication \n", + "ff656d45-d92f-4977-8077-7bcff1c29679 0 False [] \n", + " 1 False [] \n", + " 2 False [] \n", + "\n", + " num_total_samples \\\n", + "uuid replication \n", + "ff656d45-d92f-4977-8077-7bcff1c29679 0 25 \n", + " 1 25 \n", + " 2 25 \n", + "\n", + " sample_time in_group_mean \\\n", + "uuid replication \n", + "ff656d45-d92f-4977-8077-7bcff1c29679 0 0.000244 56.078323 \n", + " 1 0.000172 34.224435 \n", + " 2 0.000172 43.069929 \n", + "\n", + " in_group_var \n", + "uuid replication \n", + "ff656d45-d92f-4977-8077-7bcff1c29679 0 1044.309123 \n", + " 1 314.706175 \n", + " 2 1661.110909 \n", + "\n", + "[3 rows x 34 columns]" + ] + }, + "execution_count": 12, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "filter_dgp = \"normal\" # filter by DGP\n", + "results_df = preprocess_results_df(splits_df, filter_dgp)\n", + "results_df.head(3)" + ] + }, + { + "cell_type": "code", + "execution_count": 13, + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/Users/patrick/Projects/mis-dro-code/mis_dro/results.py:66: FutureWarning: The provided callable is currently using SeriesGroupBy.mean. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"mean\" instead.\n", + " agg_df = gb.agg(\n", + "/Users/patrick/Projects/mis-dro-code/mis_dro/results.py:66: FutureWarning: The provided callable is currently using SeriesGroupBy.std. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"std\" instead.\n", + " agg_df = gb.agg(\n" + ] + }, + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
out_of_sample_meanout_of_sample_varout_of_sample_stdsum_of_in_group_varvar_of_in_group_meanmean_solve_timestd_solve_timemean_sample_timestd_sample_time
algorithmnum_total_samples
kl_dro_bas2538.037242851.67984329.183554833.33314918.3466930.0235920.0006380.0000630.000005
10037.156579788.43294628.079048773.22917815.2037680.0349240.0062910.0000690.000009
90036.769357773.15707127.805702758.66745414.4896160.4108340.0140590.0000920.000008
kl_pp2537.787318850.78353029.168194828.92386221.8596680.0243550.0062700.0001270.000010
10037.007467796.46584628.221726780.54979815.9160470.0362350.0010500.0001350.000009
90036.686048800.69933028.296631785.51092615.1884030.4327290.0170430.0001840.000009
\n", + "
" + ], + "text/plain": [ + " out_of_sample_mean out_of_sample_var \\\n", + "algorithm num_total_samples \n", + "kl_dro_bas 25 38.037242 851.679843 \n", + " 100 37.156579 788.432946 \n", + " 900 36.769357 773.157071 \n", + "kl_pp 25 37.787318 850.783530 \n", + " 100 37.007467 796.465846 \n", + " 900 36.686048 800.699330 \n", + "\n", + " out_of_sample_std sum_of_in_group_var \\\n", + "algorithm num_total_samples \n", + "kl_dro_bas 25 29.183554 833.333149 \n", + " 100 28.079048 773.229178 \n", + " 900 27.805702 758.667454 \n", + "kl_pp 25 29.168194 828.923862 \n", + " 100 28.221726 780.549798 \n", + " 900 28.296631 785.510926 \n", + "\n", + " var_of_in_group_mean mean_solve_time \\\n", + "algorithm num_total_samples \n", + "kl_dro_bas 25 18.346693 0.023592 \n", + " 100 15.203768 0.034924 \n", + " 900 14.489616 0.410834 \n", + "kl_pp 25 21.859668 0.024355 \n", + " 100 15.916047 0.036235 \n", + " 900 15.188403 0.432729 \n", + "\n", + " std_solve_time mean_sample_time \\\n", + "algorithm num_total_samples \n", + "kl_dro_bas 25 0.000638 0.000063 \n", + " 100 0.006291 0.000069 \n", + " 900 0.014059 0.000092 \n", + "kl_pp 25 0.006270 0.000127 \n", + " 100 0.001050 0.000135 \n", + " 900 0.017043 0.000184 \n", + "\n", + " std_sample_time \n", + "algorithm num_total_samples \n", + "kl_dro_bas 25 0.000005 \n", + " 100 0.000009 \n", + " 900 0.000008 \n", + "kl_pp 25 0.000010 \n", + " 100 0.000009 \n", + " 900 0.000009 " + ] + }, + "execution_count": 13, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "cv_df = preprocess_results_df(cv_df, filter_dgp, dataset=\"newsvendor\")\n", + "cv_agg_df = get_agg_df(cv_df, [\"algorithm\", \"num_total_samples\"])\n", + "cv_agg_df" + ] + }, + { + "cell_type": "code", + "execution_count": 14, + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/Users/patrick/Projects/mis-dro-code/mis_dro/results.py:66: FutureWarning: The provided callable is currently using SeriesGroupBy.mean. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"mean\" instead.\n", + " agg_df = gb.agg(\n", + "/Users/patrick/Projects/mis-dro-code/mis_dro/results.py:66: FutureWarning: The provided callable is currently using SeriesGroupBy.std. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"std\" instead.\n", + " agg_df = gb.agg(\n" + ] + }, + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
out_of_sample_meanout_of_sample_varout_of_sample_stdsum_of_in_group_varvar_of_in_group_meanmean_solve_timestd_solve_timemean_sample_timestd_sample_time
algorithmdgpepsiloninferencenum_total_samplesnum_observations
kl_dro_basnormal0.02bayes2510037.609382835.51933328.9053518082.01828992.9147260.0293370.0037140.0000890.000007
10010036.930514782.77582227.9781317607.49329483.7886780.0427530.0037860.0000910.000007
90010036.609835775.18355727.8421187544.29743282.0077880.5026930.0172100.0001090.000008
0.03bayes2510037.622189825.11394128.7247977984.01421591.5155170.0285320.0038770.0000880.000008
10010036.984554773.60072627.8136797518.15117682.8223810.0428400.0039180.0000930.000007
.............................................
kl_ppnormal2.50bayes10010043.901629763.69487427.6350306957.543369124.2132300.0418480.0039080.0001760.000008
90010047.529628757.97006727.5312566963.104374118.0055070.4548650.0142470.0002200.000010
3.00bayes2510040.319151807.70743128.4201947617.555107107.6893950.0257370.0043740.0001610.000014
10010044.045588768.04816627.7136826977.819744126.6934070.0427090.0038830.0001740.000008
90010048.592314775.15926027.8416827036.918602128.3698280.4675190.0138530.0002260.000011
\n", + "

156 rows × 9 columns

\n", + "
" + ], + "text/plain": [ + " out_of_sample_mean \\\n", + "algorithm dgp epsilon inference num_total_samples num_observations \n", + "kl_dro_bas normal 0.02 bayes 25 100 37.609382 \n", + " 100 100 36.930514 \n", + " 900 100 36.609835 \n", + " 0.03 bayes 25 100 37.622189 \n", + " 100 100 36.984554 \n", + "... ... \n", + "kl_pp normal 2.50 bayes 100 100 43.901629 \n", + " 900 100 47.529628 \n", + " 3.00 bayes 25 100 40.319151 \n", + " 100 100 44.045588 \n", + " 900 100 48.592314 \n", + "\n", + " out_of_sample_var \\\n", + "algorithm dgp epsilon inference num_total_samples num_observations \n", + "kl_dro_bas normal 0.02 bayes 25 100 835.519333 \n", + " 100 100 782.775822 \n", + " 900 100 775.183557 \n", + " 0.03 bayes 25 100 825.113941 \n", + " 100 100 773.600726 \n", + "... ... \n", + "kl_pp normal 2.50 bayes 100 100 763.694874 \n", + " 900 100 757.970067 \n", + " 3.00 bayes 25 100 807.707431 \n", + " 100 100 768.048166 \n", + " 900 100 775.159260 \n", + "\n", + " out_of_sample_std \\\n", + "algorithm dgp epsilon inference num_total_samples num_observations \n", + "kl_dro_bas normal 0.02 bayes 25 100 28.905351 \n", + " 100 100 27.978131 \n", + " 900 100 27.842118 \n", + " 0.03 bayes 25 100 28.724797 \n", + " 100 100 27.813679 \n", + "... ... \n", + "kl_pp normal 2.50 bayes 100 100 27.635030 \n", + " 900 100 27.531256 \n", + " 3.00 bayes 25 100 28.420194 \n", + " 100 100 27.713682 \n", + " 900 100 27.841682 \n", + "\n", + " sum_of_in_group_var \\\n", + "algorithm dgp epsilon inference num_total_samples num_observations \n", + "kl_dro_bas normal 0.02 bayes 25 100 8082.018289 \n", + " 100 100 7607.493294 \n", + " 900 100 7544.297432 \n", + " 0.03 bayes 25 100 7984.014215 \n", + " 100 100 7518.151176 \n", + "... ... \n", + "kl_pp normal 2.50 bayes 100 100 6957.543369 \n", + " 900 100 6963.104374 \n", + " 3.00 bayes 25 100 7617.555107 \n", + " 100 100 6977.819744 \n", + " 900 100 7036.918602 \n", + "\n", + " var_of_in_group_mean \\\n", + "algorithm dgp epsilon inference num_total_samples num_observations \n", + "kl_dro_bas normal 0.02 bayes 25 100 92.914726 \n", + " 100 100 83.788678 \n", + " 900 100 82.007788 \n", + " 0.03 bayes 25 100 91.515517 \n", + " 100 100 82.822381 \n", + "... ... \n", + "kl_pp normal 2.50 bayes 100 100 124.213230 \n", + " 900 100 118.005507 \n", + " 3.00 bayes 25 100 107.689395 \n", + " 100 100 126.693407 \n", + " 900 100 128.369828 \n", + "\n", + " mean_solve_time \\\n", + "algorithm dgp epsilon inference num_total_samples num_observations \n", + "kl_dro_bas normal 0.02 bayes 25 100 0.029337 \n", + " 100 100 0.042753 \n", + " 900 100 0.502693 \n", + " 0.03 bayes 25 100 0.028532 \n", + " 100 100 0.042840 \n", + "... ... \n", + "kl_pp normal 2.50 bayes 100 100 0.041848 \n", + " 900 100 0.454865 \n", + " 3.00 bayes 25 100 0.025737 \n", + " 100 100 0.042709 \n", + " 900 100 0.467519 \n", + "\n", + " std_solve_time \\\n", + "algorithm dgp epsilon inference num_total_samples num_observations \n", + "kl_dro_bas normal 0.02 bayes 25 100 0.003714 \n", + " 100 100 0.003786 \n", + " 900 100 0.017210 \n", + " 0.03 bayes 25 100 0.003877 \n", + " 100 100 0.003918 \n", + "... ... \n", + "kl_pp normal 2.50 bayes 100 100 0.003908 \n", + " 900 100 0.014247 \n", + " 3.00 bayes 25 100 0.004374 \n", + " 100 100 0.003883 \n", + " 900 100 0.013853 \n", + "\n", + " mean_sample_time \\\n", + "algorithm dgp epsilon inference num_total_samples num_observations \n", + "kl_dro_bas normal 0.02 bayes 25 100 0.000089 \n", + " 100 100 0.000091 \n", + " 900 100 0.000109 \n", + " 0.03 bayes 25 100 0.000088 \n", + " 100 100 0.000093 \n", + "... ... \n", + "kl_pp normal 2.50 bayes 100 100 0.000176 \n", + " 900 100 0.000220 \n", + " 3.00 bayes 25 100 0.000161 \n", + " 100 100 0.000174 \n", + " 900 100 0.000226 \n", + "\n", + " std_sample_time \n", + "algorithm dgp epsilon inference num_total_samples num_observations \n", + "kl_dro_bas normal 0.02 bayes 25 100 0.000007 \n", + " 100 100 0.000007 \n", + " 900 100 0.000008 \n", + " 0.03 bayes 25 100 0.000008 \n", + " 100 100 0.000007 \n", + "... ... \n", + "kl_pp normal 2.50 bayes 100 100 0.000008 \n", + " 900 100 0.000010 \n", + " 3.00 bayes 25 100 0.000014 \n", + " 100 100 0.000008 \n", + " 900 100 0.000011 \n", + "\n", + "[156 rows x 9 columns]" + ] + }, + "execution_count": 14, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "agg_df = get_agg_df(results_df, [\"algorithm\", \"dgp\", \"epsilon\", \"inference\", \"num_total_samples\", \"num_observations\"])\n", + "agg_df" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Plot mean-variance trade-off\n", + "\n", + "For each DGP, we plot the out-of-sample mean $\\hat{\\mu}_M(\\epsilon)$ and variance $\\hat{\\sigma}_M(\\epsilon)$ of Bayesian DRO with different posteriors." + ] + }, + { + "cell_type": "code", + "execution_count": 26, + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAABNQAAAGUCAYAAAD9KH9CAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8g+/7EAAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOzdeVxb1503/s/Vxg5X4AXb2DHCjp2ljS2gbfbUFkm6TpsI3OWZX5upkZIuaSbTILuztU+nJaKZp0+f6UwjuU23TFsj2Z3pNGkb5OyJ2wCykzqbbQkv2MZgxJUAofXe3x/yvUZIAgkBEvB9+8XLcHXu1ZFA90jf+z3fwwiCIIAQQgghhBBCCCGEEJIWWa47QAghhBBCCCGEEELIYkIBNUIIIYQQQgghhBBCMkABNUIIIYQQQgghhBBCMkABNUIIIYQQQgghhBBCMkABNUIIIYQQQgghhBBCMkABNUIIIYQQQgghhBBCMkABNUIIIYQQQgghhBBCMkABNUIIIYQQQgghhBBCMkABNUIIIYQQQgghhBBCMqDIdQdI7jQ3N8PtdsPpdAIAXC4XNBrNjPsZjUZYrVawLIuGhgY0NzfDYDDMef+sVit6e3vhdrvh8Xiwa9cutLW1Je0PEHs8DQ0NAACHwwGLxQKLxZLWY5ordXV10Gq1sNlsae/jdrul3wUAjIyMzElfmpqa4PF4pN+vzWaDXq9PaNfc3AyHwwGO46TfaVdX15z0IR9Mfn45joMgCBntP/V51Gq1qKysBMuy4DgOHo8HAFL+fSbDcRysViv2798PANLfaGNjo3QMk8mEvXv3gmXZtPsiMhqNSX/XhKSS7+OBqKOjA93d3dOeY+12O7q6ulBXV4fh4WEAgNlszrrtfMuX8YPGjhgaOwiJl+/jhPjev7KyEgDAsuycnPtpnEhE40QMjRMEACCQZc3lcgk6nU4AIHR1dc3YvqurS9Dr9QIAobe3d9761dbWFtcfl8slsCwraLXahLZif6Z+2Wy2eetfMiMjIwIAQaPRJL19pudLfBxzTTwuy7LCyMhI0jYjIyOCRqMRXC7XnN9/vsj2+WVZNuX+IyMjgl6vF1iWnfH3bLPZBJZlhba2toTfh/j6stlsAoCUv6/p+tLb2yuwLCvodLoZHxMhk+XreGAwGAS9Xi+0tbUJAAS9Xp+yrcViSfjbt1gsSceOTNrOt3wcP2jsiKGxg5Ar8nWcaGtrSzh3t7W1Jf17pnFi7tA4EUPjxPJGAbVlzmKxSC8si8WSVnudTiewLDtvfert7RXMZnPC9q6uLgGA0NbWFrdd/KCl0+kEvV4vmM3mlCeJ+Tbd/RoMhmn3FT8szrW2tjbBYrHM+EF0pv4tdmazOavnV6PRzLi/wWAQAKR80yD+jqd7EzoyMiINZqn+nmbqi/iaXuq/UzK38nE8mIpl2ZTnMfHDRrI3nBqNJu4xZdJ2oeTb+EFjRwyNHYRckY/jhPj5INn5fGo/aZygcWI+0DixvFFAbZkTg1PJAlVTWSwWaXCZ7qSZLb1en/JkkSxqvlhe0DNdzcr2ZJyK+HsVryimytyb6fe/2C3EYCcIQsorN+IbvmTB4lRtZzvYCYIgZWoSkq58HA+mmi6gJr6pT8ZgMMSdgzNpmw9yMX7Q2BFDYwchV+TjOCGeo5LRarWzPvfTODEzGidiaJxY3mhRAgIgVmdAnDOdjNvthkajkebfNzU1zVtfnE4n6urqpPuaTJwHnuy2fOZwOHLeZ4vFAgBobW2NmwtP5tbevXvhcDjgcDjitjc3NwNAWjUQdDpdXF2DbOT6744sPvk0HmTCZrOlrOdTV1cX95gyaZtruR4/aOxYGDR2kMUkn8aJnp6elH/3DQ0NcDqd0rmLxon5QePEwqBxIj9RQG0Z4zgOdXV1ABA36CVjt9uh0+mkF7BOp5u3fokvdLGoaLrcbjfsdnvCSSbXxIKVuabRaGA2m8FxHEwmU667s2SJr43JBWvtdjs4jsuoiGdLS8us+yC+mWFZdkEX5SCLV76OB5mY7kOVuF3scyZtcykfxg8aOxYGjR0k3+XrOMFxXMrPDOL5XOwrjRPzg8aJhUHjRH6igNoy5nA4pBfmdAOj3W6XVuPp6uqa9xeQxWLByMhI0kHM7XYn3L/H40FHR4f0eDQaDerr6zMa5KxWK+rr68EwDNRqtTQwNTU1Qa1Wg2EY1NfXS1ejTCYTGIaRtnMch/r6etTV1YFhGOm4drtdGlg4jkNzc7P0Nd0VHKvVio6ODphMJjQ1Nc3ZVbC2tjZotVpYrdaMnx+TyYSOjg6pX5P773Q6UV9fD7VajaamJmlAFR8rEPvdiW3EN2RWq1U69uTn1+FwSPfV1NQEq9WatF/i/Yhfk1cxyhXxb3Py8yuuaJTJ68Ziscz6ClJ7ezsAYN++fbPanyw/+ToeZGK6D1WT22TadibLYfygsWP+0dhB8l2+jhMajUZaDXEqsY/i/zROXOkPjRM0TswGjRNJ5HrOKcmdyfPZxbnfU+dTj4yMxM2HxwLXy5ksVRFUg8GQUDC0t7d32sKNqSSbmy7WUJj63LS1tSXMVRcLRk5lMBhmLMgq/g6mLqqQ7bz8qXULXC5X0lWCUtU30Ol0Cbf19vYKGo0m4XnX6XSCVquV2osFMic/HnGlmlSPs6urK2GFV3H7VHq9Pu4Y4u89WQHZhapvIAiJtQXE2hLp1DbIpi8ul0toa2sTWJZd8FVuyeK2WMaD6WqoAUi58pQ4foj9z6RtJn1bSuMHjR3xx58tGjvIUpGv40Sqc6cgXPmbFz870DhB4wSNE7PrC40TqVGGGgGQujaZ1WqVUkjFyH6qOggmkwl1dXWz+prpygjHcWhtbYVer5eueoksFgu0Wm3cNq1WC5ZlM047NhgMCVdVxKtxya62TJ2rLl4VyYZGo4m7aiA+trlKKxfTst1uN4xG47RtTSYTenp64lKLxT7p9fqEFHONRgOn0yn9jZjNZrhcrrjHo9FowHFcysdpsVjipgaI7Ww2W9x9OZ1O2O12dHZ2xh1Dp9OhtbV15idiAUxOi54PRqMx7uqZ3W5HY2MjRkZGMkr9JmSyfB8PZkPMXkiVxTDbtpMt9fGDxo6FQ2MHyXf5NE6YzWawLAu73R53fKfTKfUznewdGidonBDROEHjRCYUue4AyQ2O41BVVSX9PHlgnHxinXzSmakOgtlsTjgpzpXW1lY0NDQknPCm09DQkDC4zmTXrl3o6OiA3W6XThZiHQiLxbIgJ5Cpz6/4u5nLIp9tbW2wWCywWq0wGo0JAUlRR0dHygKXyZ4r0dTBajLx7y7V42xsbEy4r8rKyoQ3LRqNBhqNJiEdX6PRoKenJ2mfF8Lk35M4yImPbXh4OOV+VqtVStsWH5PH40FjY2PK34FYBJaQbCy28WA2xNdlOrU5M2k72XIYP2jsmD80dpB8ls/jBMuyOHToEFpbW8GyrFS7Teybw+FIK6BG4wQ3J8encWL+0DiRnyhDbZmaOuglu9I0eZAEclcvp6OjAxzHSSeCdIknmkwGCK1WC41GE3cSGR4ehslkgsPhkI7lcDiwa9eujPqTrvm60jCVGJxMVchUvPo3+Q3UZOLfQXd3d9z2dPufql26f18sy8LlckGv14PjOOlqktvtzukKQ+JraPJrR7yaNl3mjcFggM1mw759+9DZ2Qmr1Ypdu3altWIPIdlYTOPBdFiWTZktIL7RFM87mbRN13IZP2jsmB80dpB8lu/jhFarRW9vL4BY8KCyshIGg0Hqn9gHGidonEh3fxonSLoooLZMdXd3x70YxROPy+UCEBuMpk6tnDqYLgS73Y7u7u6UwTSj0Qi1Wj3tMTJNxzYajdKg5nQ60djYKK08Khau7OrqSnnFJV25Lmyp1WrR1tYGt9s9qxV5xOd16sCS6dW6bDgcDjQ1NaG1tRUejwc6nS7r38tc9AlAXKq7uFjG5DdLqUx+85nrx0KWh8UyHsykoaFhxteX2OdM2mZiOYwfNHbMX58AGjtIflos44ROp4PBYJD66nQ64/pA48TCoHFi/voE0DiRbyigRiTiij2Taw6IZqqDAMx9zRyHw4Gurq6EaZ7iFQIgtqR1qpNrshVB0yG+IbBardi/f7+UatzS0oL9+/cnpL3PVj6k0prNZmg0GnR0dCT8DsQTrfhmaSrxpF1fXz+vfUzFbrejqakJJpMJNptNejOSa2JNv2S1/gBk9MZiId84EDJZvo0H6ZhuBa6pmROZtM3Echk/aOyYezR2kMUmX8YJcRXIqecit9sNt9sdN62UxomFQ+PE3KNxIj9RQG0ZcrvdSU/U4rzwZFeUZqqDAFwpDjmbr6mDktvtRldXV9LBoKurSxq4d+3aJaV4TyZe9Zl6wkmHWH9h6n0bjUY4nU6YTKaMaxywLJtw1WAuBst0zHS1QgxYJnvTYDAY4gpyTrZ//36wLDur53gumEwm6HS6hL/JyY/D4XDMS4Hz6frkdruTLiWt0+lgNpthtVqnre0n/u0SshAWw3iQrpaWlrj+Te3z5Cu6mbTNxFIaP2jsoLGDECD/x4menp6kf/tmsxl6vT6uLY0Tc4vGCRonCAXUliWTyZQ0yi6ujJLs5LV//36pzXzjOA5NTU1wu91obm6O+2pqaoorBtnW1obW1taEE3pzczO0Wu2si52KV6UmX1kT6x709PRk/DxMnd/ucDgSPjSmKiYpPrZMp66K+86UApzsSofIYrFAo9EkvGlwOBywWq04dOhQwj4zpZjP9DhTmXo7y7IJ9+V2u+OKq3IcJ/2tT1esMx3TPf8cx6G5uRl2u33agEBbWxtsNhtaW1thNBoTHhN3eTVbm8027ZUwsS+5njZMFr98Hw+mEl/XybAsC7PZnHCFtqOjQ5oGNJu2mVoK4weNHTR2ECLK93FCp9MlBM7EFSSnznKhcSIRjRM0TiTrC40TGRDIstHW1iZoNBoBgMCyrKDT6eJut1gsgs1mk34eGRkR9Hq9oNPpBAACAEGn0wlms3le+6nVaqX7S/Y1td/iYzMYDFJ/LRbLnPRjKrPZHPcciVwul6DX6wWWZQUAglarTeiDxWIRtFqt0NbWFvccJtu3ra1NEARB0Ov1Cb+zkZGRtPqv0+mkY7IsK+j1+hnbp2I2mwWDwSC0tbVJz/XkfvT29sbdn1arTbi/TB6nXq8XRkZGBIvFkvD319XVJQhC7O/TYDAIOp1Oek7F301bW5ug1+sFi8Uy7f2mQ6/Xx/1Nio9N/NLpdIJWq83odTEyMiKYzWZBq9VKxxOfW/F5dblcM/ZFo9EIer0+aVtCprNYxgNBiJ1/Jp8jJr8Ok52PbTabYDAYBLPZLL2uUsmkbSYW8/hBYweNHYQIwuIaJ7q6uuJeWzO9VmmcoHGCxonp+0LjRPoYQRCEuQnNEUIIIYQQQgghhBCy9NGUT0IIIYQQQgghhBBCMkABNUIIIYQQQgghhBBCMkABNUIIIYQQQgghhBBCMkABNUIIIYQQQgghhBBCMkABNUIIIYQQQgghhBBCMkABNUIIIYQQQgghhBBCMqDIdQdyied5nD9/HmVlZWAYJtfdIYQQMgcEQcDo6CjWrl0LmWz2141ojCCEkKVnrsYIgMYJQghZijIZJ5Z1QO38+fNYv359rrtBCCFkHpw9exY1NTWz3p/GCEIIWbqyHSMAGicIIWQpS2ecWNYBtbKyMgCxJ6q8vDzHvSGEEDIXfD4f1q9fL53jZ4vGCEIIWXrmaowAaJwghJClKJNxYlkH1MTU7PLychoECSFkicl2+g2NEYQQsnTNxRRNGicIIWTpSmecoEUJCCGEEEIIIYQQQgjJAAXUCCGEEEIIIYQQQgjJAAXUCCGEEEIIIYQQQgjJAAXUCCGEEEIIIYQQQgjJAAXUCCGEEEIIIYQQQgjJAAXUCCGEEEIIIYQQQgjJgCLXHSCEEEIIIYQQQkhMOBxGNBrNdTcIWTKUSiXkcvmcH5cCaoQQQgghhBBCSI75fD5cunQJwWAw110hZElhGAYVFRWorq4GwzBzdlwKqBFCCCGEEEJInjs6eAQ/OfZj3Hf9F7Bt1fZcd4fMMZ/Ph3PnzqG0tBQrVqyAUqmc0w/+hCxXgiBgfHwcQ0NDKCoqAsuyc3ZsCqgRQgghhBBCSJ4a9A/CF/Liybd/jj5fH558++coVZWiXFWBVcWrct09MkcuXbqE0tJS1NTUUCCNkDlWVFSEYDCIwcFBVFRUzNlrjAJqhBBCCCGEEJKndj9zn/T9PZvvxcETB/Dw8w8BAH77iady1Csyl8LhMILBIFasWEHBNELmSXl5OXw+H6LRKBSKuQmF0SqfhBBCCCGEEJKnHq7/mvT9HTU7km4ni5u4AIFSqcxxTwhZusQgWiQSmbtjZnuAU6dOwe12w+12AwA0Gg0qKyuh0WhQXl6edQcJIYQQQgghZLm6veYOPOX+Hd4deQcPPvclAMAW9VbcXnNHbjtG5hxlpxEyf+bj9TWrgNqzzz6Lxx9/HA6HA1VVVaitrZUKu3EcB4/HA7fbjbq6OhiNRrS0tFBwjRBCCCGEEEIyFBEikMvk2KLeip0bdDh0xgG5TI6IEIGSoYwmQgjJlYwCan19fTAajVKgrLOzc9r2Xq8XDocDO3bsQFNTE9rb27PqLCGEEEIIIYQsJ0qZEl+84ctYVbwKhYpC3LXx7lgwTUbBNEKWGrvdDrfbjba2tlx3haQh7YDaoUOHYLfbYbPZUFFRkdY+FRUVuPfee3HvvffiyJEj2LVrF/bt20fZaoQQQgiZE0cHj+Anx36M+67/Arat2p7r7hBCyLz41p++gaGJIWxiN+O7t/0rBdMy4A/7cWniElYUrUCxsjjX3SFTcBwHk8kEh8MBt9sNnU4HjUYDlmXBcRzcbjc0Gg3MZrM0K266/UQejwdGoxE6nW7a+3c6nbBYLGBZFlVVVRgeHgYA7N27N+7+Mn0MWq1Wus3j8aCysjLhMSRjsVjSCqhN7jfHcWBZFnv37oXJZILFYkmr3yR7aQXU+vr6wHEcfvjDH876jrZv3w6r1Qqr1YqvfY0KaBJCCCFk9gb9g/CFvHjy7Z+jz9eHJ9/+OUpVpShXVWBV8apcd48QQubMWGgMF/0XAQAyhqE6W2kK82HwfBSewDBCfBCewDDkjAwymZwCknmEZVlYLBZYrVYYjUZYLJa4wBgAWK1W1NfXw2azSYGqmfZzu91oamqCTqdLGWAymUxwu93Yt29fXKDL6XRi586dMJvNMwbkJvfFbrejubkZNpstIXBmNBpRW1uLvr6+lEE1MYAofk19HkQOhwNmsxldXV1x+7a2tsLpdM7YXzJ30lrls7a2Fvfee2/Wd1ZRUUHBNEIIIYRkbfcz9+Hh5x/C8ZHjuHvj3Tg+chwPP/8Qdj9zX667RgghcyoYDWLnhibUltfiavWWXHdn0TjtO4UT3AkMTQyhQlWBQDSAs2Nncdp3KtddI0lUVlamvM1gMMBkMmHnzp1p76fRaGA0GmG1WuFwOBJu7+jokGbgTQ1wabVamM1mNDU1SYsvZstsNkuZbKl0dnbCZrMBwLRZZiaTCWazOW4by7LYt2/fnPSVpC+tgBohhBBCSD55uP7KBbqx0HjS7YQQshRUFVXhq9qH8P0dP8Du9xhy3Z1FY3VxNQoVhVDKlJiITCDKR6Xty9nRwSP4ouN+HB08kuuuZMRgMKCysnLagNRUYobX1KCY2+1OGpSaTKfTQa/Xo7m5eXYdnkIM2nk8npRtent7odVqodPpYLfbU7Zzu91Jj8OybFoZdWTuzGtA7YEHHpjPwxNCCCFkmbq95g5sUW8FALx8/iUAwBb1Vtxec0cOe0UIISRflKnKUCgvRKGiECE+BLlMjkJ5IcpUZbnu2qwEIgEEIgEIgiBtC/NhBCIBhKPhpG15gZe2nR87j7cuvYmfvvkT9I+dxc/e/ClOcidw1ncmoW2EjyAQCSAUDcUdN3j5uFEhKm0TA5ULQafToaOjI+32XV1dYFkWLS0tcdvFQJper592/127dsHpdCbNcMuUGCAzGo1Jb+c4DnV1dQCA5uZmuN3ulNM3GxoaYDQak2bPZRJwJNnLaJXPZA4ePIju7u6E7RzHobOzM6u6a4QQQgghyUSECOQyObaot2LnBh0OnXFALpPHVr5jqDYOIYQsd7HAE3M5iFaO0ZAPAANBEBZlHbqW38VKMP3iQ79ERUFskcDfnDiAJ9/+Be686i58efuDUtu//v1nEIwGsa/pCawuWQ0AuN/RGnc8l/ckHn7+IennH+z4D2wovwoAcOiMA/9+9N/w/uoP4O8/8I9Smy8degCDE4P419u/h83qqwEAL517EXes/+DcP+AkxIDTdPXFxNvF4v69vb0JUzp7enrSWnBAvI+urq5ZZ36JcRGz2QybzZbyOFarVQrwtbS0wGg0Yv/+/VLNuMksFgvq6upQV1cHnU4n1YrTarXTPi9k7mUVUNuzZw+sVisaGhoS/iA5jsvm0IQQQgghSfECD6VMiW/d/G0oGAUYhsFdG++OBdOo0DQhZAlZrMGffMAwDNaVrpOev4qCimX9fN5e80G80P9cwvYieREmohM56FHmxJhDsoCa3W6Xbu/q6gLHcUkXOBD3n65m21SZxjasVmtcfMTlckGv108b7HK5XNLt4tRNq9WadFqqRqOBy+WC2WxGZ2enlEGn0WjQ1dWV9H46OjpgMplgMBhQX18PjuPQ3d2NpqYmGAyGWbdd7rLOUJtuDvCePXuyPTwhhBBCSJzfnDgA56AT+qubsW3ldgCxD06UmUYIWWpOcMfxL3/639BU1OHOq+7CTetuznWXFoXzY+chZ+RQF6qhkquk7Ys5mNb50QMAgAJ5gbTtk5vvxcfrPgE5I49r+4sP/RIA4h77g9u/iv7RM3B5XdK2TexmfPvmdjAME9d25wYdbq+5AzImvkLUv+/8IQQASvmV8fbWdbdl/+DSJAa2kgWMJgesDAaDtDJoslU1Kysrp41jiMQ26WSzTWYwGBL2cTqdqK+vh8ViSQhKud1uKftO1NzcDIfDAafTmTRLTaPRwGKxSJl4DocDJpMJTU1NcLlcCe3FhR2mLnagVqvR0NAQdx+ZtF3usqqh1tjYOO3te/fuzebwhBBCCCFxQtEQ/tv1X/jLpTfwjVf/CRf9F3PdJUIImTcuzgUuyME52IvhwHCuu7MoBCMB+CPjGA37cGH8fK67M2cKFbF6cJODgkqZMrbwglyZtG1cQIwBGEaGTexmfPGGL2MTuxkAoJArEtoqZLFtk4NsAFBw+biTA3hyWXwwbz6JgaJ0pjUaDAZwHIf29vaE27RaLTiOmzHzTKxh1tTUBACor68HwzBxX6nqnCW7T71en7SGmt1uh8vlgslkkr56e3sBJF/tc2rtNI1GA4PBgN7eXim4NpXD4UgaCOM4LuF4mbRd7rLOUPP5fCgvL096m81mw+7du7O9C0IIIYQQAMCg/yKKFcXgghxuXHsTqkuW92pthJClLRQNokxZhtHwKDQVVBspHWE+DBlk4MGDLVDnujt5QylTwnzbdxd1qQSHwzHjQgJTJQt47d27F3a7HZ2dndNOYdy/fz80Go1U90wMcs2WGAicmnXmcrlSBs46OzsTbrNYLCmngorBwqm6urrQ0NAQt02cJju1rlsmbZe7rAJqGo0GJpMJLMsmzVazWCwUUCOEEELInKkpW49/1z2OP50/jLWl63LdHUIImVd/temT+HjdJ3BpYoiCQ2kqVZWhSFmM0dDool3Rc75MDp4ttlIJdrtdWmQgE5MzqjiOA8uy0Gq1aGtrk+qEpbo/p9OZdRBtMjG4NzmY5na7UV9fn7S90WiEw+GAw+GIC2TZ7fakATUg9hiTZZc5HI64fRwOBywWS9JFGzJpu9xlFVDbuXOn9EeZLKLq9XqzOTwhhBBCSAI5I8fN627JdTcIIWRBMAyDlcWrct2NRUXOyMEWsLnuBsnAdDXN7Ha7NA1yakBHDJglW6hAp9PB4XBIMQur1Yq2tjYAkAJGTU1NsNlscccV65F1dXXNWb0wu92eEKgS+5FsGigAKRtv6uqgbrcbRqMRZrM5rt92ux06nS7lQgxutxtWqxWVlZXSAgbJZNJ2ucs6Q62npyfl7ffff382hyeEEEIIIYQQQsgSxXEcTCaTVPfLaDRKQSwxWKbRaBKCaWJ9NI7jYDAYYLPZ0NXVFRewstlsUk2yurq6hOmKZrMZTqcT7e3tqKqqkrYPDw9nlI0l9sVutwMAWltbpaDW5LpjXV1dUh+cTidaW1vhdDrhcDhgs9nigndiv4DYqqFAbKECnU4Hg8EAi8WCjo4Oqb8cx6Guri5popPD4QDLslIwcTqZtCUAIwiCMNudjxw5gu3bt8/69lzz+XyoqKiA1+tNWQeOEELI4jJX53YaI/LL+bHzKFeVoZSm7xBCsjCX53YaJ/JLMBKASl6wKFfyDAQC6OvrQ21tLQoLC3PdHbLEmEwmOJ3OtLLMMmm72KT7Osvk3J5VhlqqYNmhQ4dQVVWV18E0QgghhCweP3z93/Gu5x3ctfFD+Mw1n0WRoijXXSKEkHn1W9d/44zvNDQVGtxWcwdKVaW57lLeigpRnBs7Bxkjh7pQjYqCilx3iZC84XQ6pZVK57ItAWQzN8lcQ0MDKioq8Nhjj83H4QkhhBCyjJzkTuD1oaMIRAP488CfoJKrct0lQgiZd3++cBjPnP4jHn/jh4gK0Vx3J6/5gj7w4BERwghEArnuDiF5geM4dHR0wOFwwOVyxS3QkE1bckVWGWoAcOrUKTidzoQighzHobu7O9vDA4gVCkyVcmi1WuFyuVKuckEIIYSQxa1CxeLDtR+B43QX7tl0L+SMPNddIoSQeSUIAk55TwEAqgqrKONqBkWKIhQrSuCPjENdSKuhEgJAqoWWTj20TNqSK7IKqB05cgT19fVSsb7KykoAsRU66urqYLPZsu6gGCWdzO12SwG0zs7OlEvdEkIIIWTxW1m8Evff8EV8autnUKwoznV3CCFk3jEMgx/d9RP0efswFhrNdXfyXqGiEGtL1yLMh6GUKXPdHULIMpFVQE3MDqutrcWRI0cAXKmr1tfXB47jsuqc2+1OmuWm0Wik1SumW2WUEEIIIUsHW8DmuguEELJgihRFuLbq2lx3Y1GhYBohZCFlVUNNq9WitrYWQCzIJS7nCgC1tbUJ00AzZbfbsWvXrqyOQQghhBBCCCGEEELIXMoqoDZ5SeKKigp0d3fj9OnT0jan0znrY9vtduj1+my6RwghhJBF7JlTf0T3wGsQBCHXXSGEkAXhPnwGBx5+Gj9u+RUOPPw03IfP5LpLeW14YhjBaDDX3SCELFNZBdQEQcCePXvQ2NgIANizZw90Oh2ee+45HDx4cNaLEnAcB4/HA41Gk033EgSDQfh8vrgvQgghBKAxIt+Mh8fxxLEf4Vt/+ib+7oW/pRXuCCE5N9/jhPvwGXQ9+iIKywoQvH0UrtAJdD36IgXVUvCH/RgJenB29AyG/EO57g4hZBnKKqDW2tqKuro6aVEAvV6P1tZW7Ny5E83NzTAajbM6rtVqnZeFBtrb21FRUSF9rV+/fs7vgxBCyOJEY0R+eaH/efgjfgCApkJDK3sSQnJuvseJI7ZjqLyaRdlHCvBs5R/x8o5n4Fk3iD/9uheD/sE5va+lwBfySt8XKgpz2BNCyHKVVUANiAXVWltbpZ/b2trA8zyi0Sh27NiR8fEcDgd0Ol223Upq79698Hq90tfZs2fn5X4IIYQsPjRG5Jfq4mqsKlqFDWUb8MnN9+a6O4QQMu/jBNfvhXvYhbe/fQpNv/4kSr3lGFjbj5F+L3Y/c9+c3tdi5w/7EYgEUaIoRZGiGKXK0lx3iRCyDGUdUDt16hQeeOABNDY24tlnnwUAHDp0CAcPHpzV8ZxOJ7RabbbdSqqgoADl5eVxX4QQQghAY0S+GPQP4iR3Ar9850kMTgyiUFGIiYifsjMIITk33+MEW1OBquBK6edbtt6ClefWYKzCi4frvzan97VYhfkwgpEAPIFhRIQwokIEKwqrEBEiue4aIWQZUmSz85EjR7Bz5060tLTAYDCA4zgAwM6dO9HX14eDBw/innvuSft4VqsVLpcLJpNJ2iYubGAymVBVVYW2trZsukwIIYSQPDY5C+Oezffi4IkDePj5hwAAv/3EUznqFSGEzL/tzdej69EX4V/vw3n2LCp+VomVF9bgwsfcuL3mjlx3Ly+c9p2SvlcXqDESHMHZsVim4CZ2c456RQhZrrIKqFmtVng8HunnyVlptbW1cDgcGR0vWd00q9UKh8MBs9k8+44SQgghZFF4uP5r+D+9jwEA7qjZgYMnDkjbCSFkKdPcuAE72m7G73/WBc3bW1G4RoXTHzuBwJYxRIQIlIwy113MudXF1bjoHwAAlKrKMBIckbYTQshCyyqgNtPUTIZhsjk8AEhZb9PdPlMbQgghhCwOx0felbIOHnzuSwCALeqtlJ1BCFkWNt9ci4033gcFowDDMBAEIRZMk1EwDQBKlaU4HvRCxshwytsHuUyOQnkhylRlue5aXgpGgyiQF8xbe0KWu6xqqHm93rifBUGI+7mnp2fWx3a73TCZTLBYLACA5uZmWK1WALEgmslkgtFohNvtRmdnJ4xGIzo6OmZ9f4QQQgjJLV/Ihz+e+gNGgiOQM3IY3/MAtqi3Qi6TU30cQsiyIAgChvxD0jmPYRgKpk0yHh5HMBLERNgPBjIUygsBMAmfQwnwx1N/wIPPfhlD/qG02g/5h/Dgs1/GH0/9YZ57NvccDgfUajXcbve83Yfb7YbRaERzc7MUexDjDxzHSd/b7XY0NzeDYRio1WoYjcakCUBim7q6OopjLGJZZaht374djY2N+PrXv47t27djZGQEp06dgtPphMlkgs1mm/WxNRoNzGZz0qmeLMtK28WAGyGEEEIWtxMjx6XvP6r5GD5S91F8WPMRys4ghCwLI2e9ePfVk/hp/4/hXeVBwzX1+LuGR3LdrbwSjAaxsngVGAZYU7IWJcoSCIIwJzOjlpJgNIiDJw7gwvh5/P3Le/DtWx7FyuKVKdsP+Yfw9y/vwYB/AAdPHMAd6z+4qDLVKisrodFoUFlZOS/H7+jowP79+2E2m6HT6aTtYqLP5IUV9Xo99Ho96uvrwXFcyniFzWZDU1MTbDYbWJadl36T+ZdVQG3nzp0wmUz4whe+EJetxrIsrFYrtm3blm3/CCGEELJM1K9uwE/u/jmeO3MIjdXvA3A5O4PqBhFCloGL717C6798G9txE459oBvF7y3OdZfyTlVRFcpUZRgNjaJYEXt+KJiWqEBegH+5+TtSkGy6oNrkYFp1cTX+5ebvLKpgGhArRdXb2zsvx7ZarWhvb0dfX19C4EtM9Kmrq0soh2U0GmE0GuOCbZNxHIempiYKpi1yWU35BGIRWI/Hg+7ubjz++ON45plnMDw8jHvvvXcu+kcIIYSQZaRcVY6/2vRJrC1dl+uuEELIghofHpe+V69S46ryjbnrTB5TyVWoKqqiQNoMVhavxLdveRTVxdVSUG3q9M+pwbSZMtmWG3Gap9lsnjbwZTKZEraJCy6mylCzWq1JF2Uki0tWGWqTabXahMjrqVOnsHHjxrm6C0IIIYQQQghZkjbdVgu2pgLjw37U3vhJlK0syXWXyCInBtWSZarlSzDN6XSitbUVbrcbhw4dkuqgdXV1AYgFpBwOB9xuNziOQ3d3N/bt2ycFuMSgV09PD2w2G3Q6Xcpjdnd3A0DSslLJiO1aWlqmbdfS0oL29vaE7QaDAVarNWlQzeVypZWdlu3zI3I4HHA6nWBZFr29vTAajXHxG7FePYCE2+fq+VyShHl0//33z+fhs+b1egUAgtfrzXVXCCGEzJG5OrfTGLFwwtGwwPN8rrtBCFkG5vLcTuPEwlnqY8TExITw1ltvCRMTE/Ny/MHxQaH1j38jfOw3HxZa//g3wluX3oz7eXB8cF7uNxMABIPBIIyMjEjbWJYV2trahN7eXmmbXq8XDAZD0v27urrSOqbNZkurTxqNRmBZNrMHMklvb68AQLBYLHHbu7q6Evo6k2yeH5vNJmi12rhtLMsKLpdLEARBsFgsQltbm3Sby+USAEi3z9SHdJ/PXEv3dZbJuT2rKZ8+nw8PPPAAGhsbE742b94cF+UkhBBCCEnmv13/hQef+xJ+5/4f+MP+XHeHEEJInrkwfgHnx85hLDRKK3rOwtTpn6aXHsl5ZtpULMtKX6KGhgY4HI64TKrGxkb09PQk3T/dY4qZVTNxu91ZLXQgzuKbmqHW1dUVt7hBOrJ5flpbW7F37964bS0tLXGZZQ6HQ/peo9GAZVk4nc60+pDu87kUZTXlc/fu3QBiv4ypf8AjIyMUUCOEEELItHiBxx9P/QED4xdgfeNxaFfVo1hJhbgJIYTEhPkw/JFYfblQNIwSZWmOe7Q4rSxeib+t/zuYXrqycuzf1v9dXgTTRI2NjXE/sywLjUaTsC3bY6ZLo9HA4/FkdH9TTV2cgOM4VFVVzepYs3l+nE4nOI5LKM9VX18vBfoMBoNUz43jOGlKZ7LHns3zuRRlFVBrbGzEI4+kXsqZCkUSQgghZDq+kA/qAjUGxi/ghpU3YG3p2lx3iRBCFhwf5THcP4JvHvtHVLFVeO+qG3DvZn2uu5UXInwESpkSYT6M8oJy+ow5S0P+IXyv91/jtn2v91/zJkMtlVwGbHQ6HaxWKziOm7Efdrsden3ia9ZgMMBoNMJiscBisaRcjKC+vj4hI6y3tzfpCqGTzdQvMVvN4XDEZdtVVlbGZajZ7Xa0t7ejoaEBRqMxq8y85SSrgNpMv7zpgm2EEEIIIWwBC/Nt38Vp3ylE+Eiuu0MIITkxOjiOgw/+Hu/BTeiv68OJTx3PdZfyRpGiCFeVb4Q/7IdKrsp1dxalqQsQ/G393+F7vf+asFABiWcymWC1WtHZ2TntipxiRlcqkxcnGB4eThpH6e3tzba7SYkZbDqdLiGbTdTR0QGLxYKurq6UbUhyWdVQ02g0OHr0aMrbp87TJYQQQghJ5qryjahjN+W6G4QQkhPjniv1I0OFQawurs5hb/JTsbIYCllW+SDLUrLVPK+pujauptrfv7wHQ/6hXHc172g0GlgsFphMJnAcl7Jdquw0kdFoBAA0NzcnTJmcbzqdLmk9NOBK3TSTyQSLxRIXTBOne3IcF1dfjcTL6oy0c+dO7N27F62trWhoaIiLtIpPfLLlYwkhhBBCCCGExBQUq3D1BzUYuzSO9916A2o3r891l8gSkCyYJmaiiQsViLfnY6Yax3EJgaxk2ybfNptjTsdgMIDjOOzcuRP79u1LmIJptVpnXGBAq9VCo9HA4XDAZrOlfd8zSff52bdvH0wmU1zQz+12x7Wb/L3b7YZGowHHcfB4PNNO/8z0+Vxqsgqo3X///ejs7ERDQwOGh4cxPDws3SY++YQQQgghU/ECD0/AgxVFK3LdFUIIybmqWjU++NBNue5GXonyUYAB5Iw8111ZlKYLponyIajmdDphsVjAcRza29vh8XjQ0tKC9vZ29PT0wO12w2QywWw2S1MT3W43jEajlDk2eX/gSmZZqmOKdcWmrr6ZSltbG/R6vVRzjGVZaWEBg8GQVp03s9k8q9Uws31+NBoN9Ho9NBoNjEYj6uvrAcRqqIkBtq6uLmk/rVaLyspK2Gw2mEwmALEsN6PROGfP51LCCFmsO3z//ffj8ccfT3n7nj178Oijj8728PPO5/OhoqICXq8X5eXlue4OIYSQOTBX53YaI+ZX78UefOvwN9FQ3YDmq3dhS+XWXHeJELIMzOW5ncaJ+TU8cQlc0IsyVRkqCyuX9HTPQCCAvr4+1NbWorCwMOvjBaNBPPjsl3Fh/HzKYNpkk4Nva0rW4v/t+AEK5AVZ94OQfJLu6yyTc3tWNdSampqmvZ1qqBFCCCEkmT+e+gN48Hht4DWMBEZy3R1CCCF5RBAE+EI+COAxGvKBAa3smYkCeQHu2Xwv1pSsTSvjTMxUW1OyFvdsvpeCaYSkKeswv8/nSxm1s9ls2L17d7Z3QQghhJAlZhO7CSe5E+AFAY3V78t1dwghJOd+9uZPUVVYiQ3lV+G9K2/IdXdyihd4lCpL4QuNokRZDLmMpn1m6q6Nd+OO9R9MOzi2snglZaYRkqGsAmoajQYmkwksyyZdrcJisVBAjRBCCCEJWrZ8CvdubsaAf4A+KBFClr2DjzyN89wI3lX34dWPvbrsA2pymRwri1ehsqgKvMDnujuLVqbBMQqmEZKZrFf55DgOLMsmLUDn9XqzOTwhhBBC8pz78BkcsR0D1+8FW1OB7c3XQ3PjhrT2lcvkWFe6bp57SAgh+e3kK6cxdNyDSqxExbAa0TP+XHdpTgXHQvCPTCAajkKulKNYXYSCUlVa+8oZOS1KQAjJW1nVUNNoNOB5Hh6PJ+lXa2vrXPWTEEIIIXnGffgMuh59EYVlBWj87A0oLCtA16Mvwn34TK67Rgghi4L78Bkc6ngJMnmsRpiySAlVZ/mSOY8Gx0LwDYxCJmdQXFkEmZyBb2AUwbFQrrtGCCFZyypDbd++fdPebjQaszk8IYQQQvLYEdsxrLuhGh/+xg4wDIP3fPwaPP2NZ3HEdixllpr9uA0vnH0OX3hPK7at2r7APSaEkPxyxHYMNdvW4MPf2IFoKIpwIIJn/88r055HFxP/yASURQpUrL1cc5stgve8D/6RiaRZamOhMZwbO4f1ZetRrCxe4N4SQkhmsspQ2749/o3w0aNHp72dEEIIIUsH1++Ff2QCBx9+Gn/+xRFEQlGs164B159Y8mHQP4jugdfw87d+itOjp/H/nP8XJ7kTGPQP5qDnhBCSH7h+L9Zr14BhGCgKFCiqKEx5Hl2MIqEIxocnwJ3zYcIbAAAoi5WIhqNx7cJ8GMFIAOfG+nF+/ByOXXoDnolhhPlwLrpNCCFpyXqVz8nq6+sRjUZnbkgIIYSQRa9iXTk8pzgIvICxYT8aP3MDzjovgK2pSGi7+5n74n6+FLiEh59/CADw2088tRDdJYSQvMPWVOCs8wKiHwjgp28+gc9f9zc47xxJeh5dlAQgNB6CqlgJmZxBUUUhwv4w5Mr4uminfacAAMOBSyhXlsEXHsWF8QvwBD3YxG7OQccJIWRmWWWoTSUIwlwejhBCCCF57Jq7NkPgY2N/6YoS/P5/P4f+oxegbb4+oe3D9V+Tvi9VlibdTgghy43m4+vRf/QCDn3rFeBVFZ77l8PoP3oBdX+1PtddmxOF5QWIRniMDY0jEo7Ce96HkD+MYnVRXLvVxdUAgGJFCaqKV0IlU6JAUSBtJ4SQfDSnATWGYebycIQQQgjJY9fdfTWa9tyGyqtYjJzlEBgN4s49t6E2Sd2f22vuwBb1VgDAWHgMALBFvRW319yxkF0mhJC88s/cHry77Q0ohlXY0nsDIqNR9Ox4Cf80sifXXZsTZatKsX77WpStLkU0GAUfFVBeXZZQP61MVYZCeSFKVaWI8GGsLqlGobwQZaqyHPWcEEJmNqdTPgkhhBCyvGhu3JBW4eyIEIFcJscW9Vbs3KDDoTMOyGVyRIQIlIxyAXpKCCH55+H6r+H3rx1Ckb8EAPDmjb0YWX1pSWXvFpSqUFBaOW2b2Ewn5nIQrRyjIR8ABoIgUNIGISRvzWlAjRYhIIQQQkgySpkS37r521AwCjAMg7s23h0LpskomEYIWb5ur7kDrwmvSz+HC0LLMnuXYRisK10nBc8qCioomEYIyXsZTfl87LHHpr29p6cnq84QQgghZOnxBr1wcSelYBoQ+/BEwTRCyHIXESIYa/Dgwkf7UNlSgpp1NVL27nISioYSHjMF0+aPP+zPdRcIWRIyCqhZLBaMjo7OV18IIYQQski88G+H8dQ/H8Ib//02wsHpP/i92P8C/vb5r+ILz9yH3ot08Y0QQkRKmRL/+Ml/xD/v/gc0f/aTMOs68K2bv70kLjiMnOHguziGwGhwxrbDgWGc9p3C2dEzCPPhBejd8uXiXPjs05+Ci3PluitzxuFwQK1Ww+12z9t9uN1uGI1GNDc3w2g0oqOjAx0dHQAAjuOk7+12O5qbm8EwDNRqNYxGIziOSzie2Kaurk7alyw+GU35dLlc0Ov1aGpqgsFgQHl5+Xz1ixCyAI4OHsFPjv0Y913/BWxbRVO2CSHpEXgBp17rR8AXxMV3L+G6j2yZtv2fLhwGAFyaGMKKohUL0UUyB/xhPy5NXMKKohUoVhbnujuELFmTg2cMwyyJupKRYASRUBSRUBQCL6CwrCBlW17gpYypCB+BgqEy3/Ppxf4XEBWieOncC6hj67I6VpSPIiLEfmdymXyOepi5yspKaDQaVFZOX6tvtjo6OrB//36YzWbodDppO8dxMJlMcDqd0Gq1AAC9Xg+9Xo/6+npwHAeLxZL0mDabDU1NTbDZbGBZdl76TeZfRmcrvV6Pzs5OeL1eWCwWuN1uNDc3Y8eOHfPVP0LIPBj0D8IX8uLJt3+OPl8fnnz75yhVlaJcVYFVxaty3T1CSJ4bG/ZDroy9cV733mrIFdMnvN9ecweUMiWGJoawoeyqhegiyUKYD4Pno/AEhhHig/AEhiFnZJDJ5Esia4aQfDYwfgEHTthRXbIGAFCiLMHdGz807T4HTtgxMH4BdewmlCkTV8W8ed0t89LX6UTDPBgZA4EXoCqa/rwhQIC6UI3x8BgK5IU01XMeCYKAl869CAB4qf8lfO7a+2b1fPMCDwCIClEIgoAoomCE2HFkTEaT4OaEVqtFb2/vvBzbarWivb0dfX19CYEvlmVhNptRV1cnBdRERqMRRqMxLtg2GcdxaGpqomDaIpdRQK2zsxMAUFFRgUceeQQAcOjQIezZswcrVqygrDVCFondz9wnfX/P5ntx8MQBPPz8QwCA337iqRz1ihCyWJStLMFnf/xJcGe94KPCjO3v3HgX7tx4F6J8lD4oLQKnfaek79UFaowER3B27CwAYBO7OUe9ImTp6nJ24akLv4OsiEH/+Fk8cdfPUKoqBQD89M0ncOCEHfdu1qfcf2D8Av546g9Jb6urqMtJQK2gVAVViRqRQAQy5fSZS3JGjsrCSlQWVl5e7ZPMF7fXjUsTQwCAoYlB9Hnd0MwiSy0UDUnfK2RyRPiotK1QUTg3nc0D4jRPi8UybeDLZDLB5YqfQmswGKR9k2WpWa1WGAyGue4yWWBZ59Pu3LkTO3fulLLWhoeH8b73vQ/33HPPXPSPEDLHokIUf6v9O3zP+a8AgDtqduDgiQMAsKSWaCeEzC+GYaDewGa0Ty6ng5D0rSpajcGJiwCAUlUZRoIjAIDVxdW57BYhS1IkGIH7mxdxDRrRvfMF3PD+bVIwDQD0m1vwmad3TRtQA5JfED1wwo6b1948531OF8MwUM6QnZZsHzI3+kfPos/bF7fttYE/Q8bIwAs8ZIwMvzl5EO+rfn9cm9qKWtSUrZ/22EqZUqp1J2PkAKLS9kw5nU60trbC7Xbj0KFDUh20rq4uALE67g6HA263GxzHobu7G/v27ZMCXGLQq6enBzabDTqdLuUxu7u7AQBmszmtvontWlpapm3X0tKC9vb2hO0GgwFWqzVpQM3lcqWVnWa329He3g6n0wmdTidNEXU4HGhubkZlZSXMZjP0ej2sVqu0X29vL4xGo5Qd53A4pOmpLpcLdrsd3d3dUjkvMjuMMA+XAVpaWnDo0CEYDIakf1j5wufzoaKiAl6vlzLryLLxi7d+hv7RflyaGMIJ7oS0fYt6Kzpue4zeyJBFb67O7TRGkOVIEAScHz8PLjCCAsWVmkeF8sIZP2ARshjM5bl9Lo41PuzHk39zEADwh//ViY9s/Rh2v6c1rs3H/+sj+N83/UvKerevnHs5IQvNxZ3ECe7EjNNFSX4IBALo6+tDbW0tCgvnJsOr/c/fxuELr2a8301rb8ae93095e28wCPMhxOyCWWMDCq5KuP7EzEMA4PBALPZLAWa1Go1DAYDdu3aJQWGxCDS1CAVwzDo6uqKq3GW6pj79u2DXj99kBoA6urq4PF4MDIyMqvH5HQ6UV9fD4vFEhe0cjgcABDX1+lwHAe1Wp3w+EwmE4xGIzQaDaxWK1wulxQEdLvdqKurg8vlgkajiTuO2WxGW1sbTCbTtHXelpp0X2eZnNvnbILz0aNH8cADD0Aul8Nut0OtVqOxsXGuDk8ImQOHz78K2/FOHL7wKs6MnsHV7BZ88YYvY4t667Jcop0Qkhn34TM48PDT+HHLr3Dg4afhPnxm2vYD4xcw5B9aoN6RuXBp4hL8YT+8QS/8YT9WFq1CobwQAENTsQiZDwyw6baNWNVYiagyCu2qxFpLJcoSuLypV2RMNqXzD6d+n5NgWnAshJGzXlxyezBy1ovgWGj69pEAokJ0gXq3vDyofQi3rLs1o31uXXcbvrL9qylvFwQBET4i/Q8IUMgUUt20bMYJlmWlL1FDQwMcDkdcDbLGxkb09CSuGJ4s2yvVMcVMtZm43e6sFjrQarXQarUJAaupgbGZsCwrZadNVlVVJQXLgCuBOgDQaDRgWRZOpzPuOJOZzeZlE0ybLxlN+fT5fAkRuh/96EewWCxwOp0QBEGaK7x9O60YSEi+EFfzvGntLShRlGA8Mo5Pb/0sPrnpHjAMg7s23o2IEKFi04SQlNyHz6Dr0RdRtqoEbE0F+DCPrkdfRNOe26C5cUPSfX79zq/w7NlD2MRuQlvjHqnINsk/4oqeckYOhgFWFa9CTdl6FCoKUVFQAUEQKIOZkHlQUlmMnX93C1zcSeD55AXdy5RlGA350j7mT998Ap+79r6ZG86x4FgIvoFRBMeCKCgrQDQUgW9gFOXVZSgoTZ65dGF8AFEhgiJFMdaWrl3gHi9tJcoSPNJggnZVPX74+r8jKkSlxQQmkzEyyBk5vnjDl7Bjgy7luT7KRy8H0WKUMiVUcpXUfi7GiakJOSzLxgWMxG3ZHjNdGo0GHo8no/ubauriBBzHoaqqalbHaW5ulgJgUxc7MBgMUhYcx3HSNNdk/U+2SAKZnYwy1EwmE4D4bDSDwQBBEPD444+D53k8/vjjFEwjJE8M+gdxkjshrebZffHPeFD7ED5S+1EpmAZcrnFBwTRCyDSO2I5h3Q3VEAQBl1weeC/4sPa91ThiO5a0fZSPonvgNQDAubFzqCpcsZDdJWkK82EEIwFpRU8BPCoLKlFZVBlXWJqCaYTk1nh4PK12A+MXMB4ej6vDtlD8IxOQK2UoKC0ABEBRoICqWAn/yETS9sFIABEhDAECAMqAnQ8Mw0B3VRO+/8EfQJHivb5CpsT3P/gD7LyqKem5nhd48AIfW9ETsaCZnJFBIVPEtZ+vcSKXq2DqdDpwHAeO42Zsa7fbk24Xg1xiICzVYgT19fVgGCbua3J2mV6vB8uy0v04HI6ELDe73Y76+nopbpMquy6brDsSL6OA2v79+7F582ZpHnBrayt6e3vR09OD1tbWmQ9ACFlQu5+5Dw8//xCOjxzHPZvvxfGR42h/7dt4qu939OGIEJIRrt+L6mtWwD8SAABUX7sKVzWsBdfvTdo+zIfxsbqPQ1NRh4bVjVDKKWifj077TuHs2FkEogGoC9QIRAPwBD3wBLK7Ik8IyUyJsiTlbaPh0bSPc+CEHdtW5ia5IRqOglFc+XipLFZCWaxENJx8SqeMkaNCxULBKFGiXPgA4HIiY2QIRYNJbwtFg9MuGhSKhhCKhsALPBQy+eXgGi8tSrCUiYGpzs7OaduJ2WCpiIsTAMDw8HDSIGFvby8EQYj7mppJ1tLSknKKZkdHB0wmE2w2GywWC2WhLZCMAmocx6GiooKy0QhZJCav2nlHzY6k2wkhJB1sTQUuvjOMz/+yBX/16J2o//R7cdZ5AWxNRdL2hYpCfGrrZ/B/P/j/8FD9wwvcW5Ku1cXV0hSgUlVZ3HZCyML4H9dv8UL/CwCSZ6KNh8enDbhN9vK5l1DH1s1p/9IlV8ohYxhU1apRvqYMhWUFCPvDkCuTB2uUciVWFq/ExoqNKFfR4j/z6fD5V8AgdjFdnFYs/s+AwavnX0m57+RZLLEVPRO3L1UajQYWi0Uq3p+K3W6fdpEDo9EIILagQjZ15o1GIxwOB6xWa8L9mUwmWCyWuCmy4nRPjuPi6quRuZNRQE2v11M2GiGLyHVV10N+eeB78LkvAYit5nl7zR057BUhZDHa3nw9+o9ewDPfeQGDxy/B+eu/oP/oBWibr59x3+XwpnuxkjEyeCY8mIhM4OxobJGJQnkhyiYF1wgh8+fY797BOx2n8e6/nQYjyFJmo6WTdXZ08AjGw+M5q1dZrC5CyB/G6MUxRMNRjA2NI+QPo1hdNOO+NHNifr107qXLU2tjnw++d8f3cV3VdQAAAQJe7n8x5b48rkz3DEVji0zIGNm0WW1zJdl0y+mmYKYzNTPdKZwig8GAvXv3YufOnXFTMEVWq3XGBQa0Wi00Gg0cDkdaq4vOdBybzZZQWw6If/xutxsajQYcx8Hj8dA0z3mS0aIEu3btmq9+EELmmCAI+Lcj35dWTnpf9fvhDXql1TyVDH3AJYSkT3PjBjTtuQ1HbMfQ/Z+vg62pwJ17bkNtigUJSP6L8lEMjl8EL/AYDfqwqmg1wnwI4oqe9AGXkPk30u9D+YAa5VCjPFqBgfELcbeLP29bNXNAbcA/MC99TFdBqQrl1WXwj0zA75mAXCmfdkECsjCG/INwe12QMTL89bWfwyc33QMZI8O3bv4OfnPiAH7x9s/h8row5B/CyuKVcfvyAo8oH4UgCIgKURQqCqWs5rkaJ5xOJywWCziOQ3t7OzweD1paWtDe3o6enh643W6YTCaYzWZ0dHTAYrHA7XbDaDRKmWOT9weuZJalOqa4Smi6K1y2tbVBr9fDbDYDiNV1ExcWMBgMadV5M5vNaa8uOh1xGupUXV1d0nOj1WpRWVkJm80mtddqtWhubgYAtLa2QqfTSY+HzB4jLOM10H0+HyoqKuD1ehNWLyVksYvwEfzszZ/it67/grpQjR/s+A+UKEtpNU+y5M3VuZ3GiNk74zuDFUUrUKwsznVXyDQifASD/osYD/tRoiyRVtijYBpZyuby3D4Xx3ru/76K48/F6i9V/1M5nhr+b1jv/LF0+0/ffALVJWtw98YPAQDGQmMwd7fj89fdhzp2U9yxfvrmEzh44gB++4mnZvmIFk4wEoBKXkDnmssCgQD6+vpQW1uLwsLCmXfIwKWJS/j3o/+GT2/9DK5Wb0m4/V3PO/j1u7/Cl7Z9BSuK4hcQ4i/XShMEAQqZAgpZLB+HxgmyGKX7Osvk3J5Rhtp3v/vdhGVXxSgwABw4cAAMw+Cee+7J5LCEkHmgkCnwhffsxgfWfAARISrVxqHMNELIbFx8dwin/tyP1VtXYs21q6bNOPhujxnnxvqxbeV2fP39/yC9ASf5RSFTYG3pOviCXhRPqs9EH5IIWTgffOgm3Pal9yM4HkJhWQGuH70WP33zCWxmr8aAfwBlqnIpmAYAY+FRnOROYDSUODW0umQNqnNY/zDgC4KP8lAWKqAoVKQ8l0T5KM6OnYWckaNcVYGqoqoF7unysqJoBf75xm+mvH1L5daUt8sYGVQyFaJCVCojA9A4QYgoo3e4BoMBzc3NcLvd2LNnj5QyKLr33nvR19eH7373u2hqasK2bdvmsq+EkFm4bsXM9Y0IIWQmp187h6MH3gQA6NpuRd3NVyVtd2H8Ak77TgEARkOjFExbBMoLki8sQQiZX+7DZ3DEdgxcvxdsTQW2N1+Puhs3JWSeTVZdsga/+kjyFQfv3vihuODbQgv4AggHIgCAqlo1GHnyoMt4eAwALpclWbaTpRYNhmGgYGgsJySZjBYlqKiogFarxcmTJ7F7925UVCS+AautrcUjjzyC7u5u+Hy+OesoISQ9zou9+OqzX8bRwSO57gohZAkZeGdQ+r5668qU7WSQ4cO1H0VVYRU+sPbGhegayZA/7McZ3xn4w/5cd4WQZct9+Ay6Hn0RhWUFaPzsDVCWKtD16ItwHz6T667NisALCAdjwTS5Ug6ZPPXHTKVchVJlKWSQoURZulBdJGmK8lEEo0FE+Wiuu0JI3sso1PzYY4/h61//elptW1tb8dhjj+FrX/varDpGCMnMoH8QLu4kHuvpQJgP48m3f45SVSnKVRVYVbwq190jhCxyuq/dioG3h+A5w6GkKnVttNUlq3H/DQ/A+N77EeEjC9hDMpMwH4Y3wOG07xRKVKXwBBjIGRlkMjnV1iRkgR2xHUPNtjW44+8/gGPDx/AE/2/4TKQVR2zHoFmEi70wMgbq9SwigZnP+0WKIhQpiqgOV54RFxsIRoMQwAMygBFivx8Zk1EeDiHLRsa5m5kU3Jyr9Q6amprQ1dUVt83tdsNsNqOurg5AbKUNg8EwJ/dHyGK0+5n74n4+PnIcDz//EAAsiuK0hJD8VqwuguamDdDcNP0HvaODR/CTYz/Gfdd/Ia1V6cjC6fO6cWH8AniBB3P5w1EgGgAAbGI357JrhCw7XL8XjZ+9Ac/+9kX8wf00VqnWwr3yXaw8XJPrrs2aQiWHQiWfto0/7MeliUu0cE0eCkVDEAQBESECOSOXfgaAQsXcLpRAyFKRUUDN5XJldPCpCxjMRkdHBxwOR9w2t9uN+vp69PX1SUvUmkwmdHR0oK2tLev7JGQx+uINX8Z/vP4DAEBlYSU8gdjr7+F6yhIlhMy/Qf8gfCEvnnz75+jz9VGWbB6qLKzCwPgAgFiR6hAfAgCszmERc0KWK7amAmedF3Dh7WG8N/h+jLJehEtCYGuWZk3DMB8Gz0fhCQwjxAfhCQxThmyeUcqUCEaDYBDLSGPASNsJIclllLuZaYCM47iM2k/ldrvR3d2dsN1sNsNgMEjBNADYu3cvTCZTVvdHyGJ218a7cTUbWwpbDKZtUW/F7TV35LBXhJDlYvcz9+Hh5x/C8ZHjuGfzvVKW7NTsWZI7K4pWYGP5RpSryuGPxOqnFcoLUXZ5FWhCyMLZ3nw9+o9egBCLa6NorBiV51Zi3YdX5LZj8+S07xTOjJ6BL+SDukCNQDSAs2NnpUVsSO7JZXIoZAooZErwAg8ZI4OMkUEumz7rkJDlLKOAmlqtxrPPPptW26NHj2Y95dNut2PXrl0J2zs7O6WpniIxuDY1m42Q5SIiRKCQK7BFvRVfvOHL2KLeCrlMjohANYwIIbMnCALe/P1xDJ0YRjTCp2z3kPZh6XtvwCt9T1my+UMQBMgYOVYXr8aakrUolBcCYOasRAchZHr9o2fxUv+LeKn/RZxbfwo1rSvAl0cQlUURLAygd8dLeIV9QWojfvWPns1116cVjfAI+IKIhFIXsV9dXI0QH8JF/0V4Ah6Mh8el7SQ/iGOBnJGjQF4g1U2jMYKQ1DKa8mkymdDS0oJnn30WZWWpr2b6fD60trbCZrPNumN2ux16vR5OpzNuO8dx4DgOGo0mYR+WZeF0OqHT6WZ9v4QsVkqZEt+6+dtQMAowDIO7Nt6NiBChNG1CSFZGL47h5cdfAwBsaFyHD/3DB5O2q624Mi4fOhu7uEVZsvmFYRisK10nFQGvKKigouCELKBfvPVzHL7wavzGe+J/HOjvxwv9z8dtu2ntzdjzvvQWhsuFkD+EscFYgKxkRTGK2aKENmWqMsguTyHsH+uHukBNGbJ5hmEYKGXKuDGBxghCppdRhlptbS2am5uxceNG/PjHP4bP54u73efz4Uc/+hFqa2uxa9cubNy4cVad4jgOHo8nadDM7Xan3K+yshLDw8Mpbw8Gg/D5fHFfhCx2B08ckGriTB4ExUGREJIeGiOSG3hrSPp+1eaqlO1CfAilylLIGTluWnMzZcnmCV7g4Ql4EBVimSNTPxjRByVC0pftOPGg9iHcsu7WjPa5dd1t+Mr2r2a0z0KLTFw5zysLkudrxDJkFShXlqGyoArlqnJQhmx+4AUe0curctMYQUhmMl7ls62tDSzLorW1VapjVllZCY/HI9VMe/zxx9Ha2jrrTlmt1lkvLjBd3bb29nZ885vfnGWvCMk/Rwad+OmbT+CXbz+Jv772c/irTZ/IdZcIWbRojEjkPnwGR3/zJmQKBspCJRh56utwW9Rb8LMPPQk55IgiCgWjoCzZPDAS8GAkOAJv0IvVxatpVT1CspDtOFGiLMEjDSZoV9Xjh6//OyJ8BAISA0oMGChkCnzxhi9hxwZdXgc1gmMhBEaDCI6FwPMCylbxUCYmqIFhGFytvhoMw0hZT5T9lB8ifCQWVBOiUMiU0lRPQsjMZvVqMRgMOHnyJHbv3g21Wg2XywW1Wo3W1lacPHkyq2Caw+GYdsrm5IUIpppp0YS9e/fC6/VKX2fP5nc9AkJmYj/eCSCWGUIp84Rkh8aIeO7DZ9D16IsoqSzG+/+/7Vi5qQrdvzgK9+EzKfdRypSQyWRStiwF03IrKkThDcbq2fFClH4fhGRpLsYJhmGgu6oJ3//gD6BglJCH5ZBFZGD4K4ElhUyJ73/wB9h5VVNeB5yCYyH4BkahKlaiciOLiupSjA6OITgWStp+8iyKyf+T3InyUfBCrD6qAEgrexJC0pNxhppIo9HAYrHMZV8AAE6nc9rstMrKSgDJM9E4jps24FZQUICCgoJsu0hIXjg6eAS+oA83rb0ZvqAPd6xPrGvkPnwGR2zHwPV7wdZUYHvz9dDcuCEHvSUk/9EYEe+I7Rhqtq3Bh7+xAwzD4D0fvwZPf+NZHLEdo/PIIuAP+3Fp4hKqCldgPDKOArkKSnliQC04FoJ/ZALRcBRypRzF6iIUlKpy0GNC8t9cjhMyRoZoOIoPPdkCABiuvog/fSi2+FuYDy2KlRX9IxNQFStRsbY8toEtgve8D/6RCTqP5LlgNChlkitkilh22uU6zFPxUR58lIcgAAwDyOQyyKbJWCdkOVnwV8J0tQasVitcLhdMJpP0JQbtTCYTOjo6wLIsWJZNmY3W1NQ0L/0mJF8M+gdxkjuBJ9/+OU6PnsaliSF8/vr7cGniUlw7MbuksKwAjZ+9AYVlBeh69MVps0sIIUTE9XuxXrsmLpNgvXYNuH5vQluqgZM/wnwYwUgAnsAwQnwQo2EfqgorUaYqT2grZpfI5AyKK4sgkzPwDYymzC4hhMydw+dfgUy48lFMYK6cRxkwePX8K7noVkai4SiUxfGBemWxEtFw6tU+Se79ZegNfPbpT+Evl/4CQRDACzyUjDJlMC0a5gEwl4NoDKLhWIAtnzgcDqjV6mnrrWfL7XbDaDSiubkZRqMRHR0d6OjoABBL7BG/t9vtaG5uBsMwUKvVMBqNSZOBxDZ1dXXSvsuduADlTNxuN5qbm6FWq3P+3KWVoeb1erFnzx6YzWaUlye+IUvXkSNHcOjQIXzta19LervBYEjYZrVa4XA4YDabpW0tLS1wuVxx7cQXD63wSZa63c/cJ31/z+Z7cfDEAXzthYcBAL/9xFPSbZRdQgjJBltTgdOvncPVOzQoKI1lZJx1XgBbU5HQ1nZ8P17ofwFb1Fugv7oFa0vXLnR3yWWnfaek79UFaowER3B2LDYtbRO7Oa4tZZcQkjsvnXsJAnhcWjOAEnkptl+3Df4Vl2JBDgh4uf9F3LtZn+tuTkuulCM4GkRReSEYWSwYE/aHIVcmZtedHzsPXoiiQF6IFUUraLpnDv3irZ8hFA3hV+88iW/f0o4IH0VIiF1IKVQUxrXlozwYGQOF6srvNBKKgo/yeZWlVllZCY1GI81mm2sdHR3Yv38/zGZzXLyB4ziYTCY4nU5otVoAgF6vh16vR319PTiOSzmrz2azoampCTabbdpZdstJT08PKisrpecyFY1GA5vNhrq6ugXqWWppvQoqKirw6KOPYvfu3fjRj36U8Z34fD7s2bNn2mBaKskilCaTCXa7PW6bxWKZlymohOSbr2x7UPr+jpod0vcP18e/trh+L1Zurkwru4QQQqba3nw9zh+7iJ/9Lzt+3PwrHHj4afQfvQBt8/UJbd/xvI2zo2fgONOVg56SyVYXV0v1cEon1dZcXVyd0DYSjEBeEP/Bl7JLCJl/Q/5BuL0uCCoBW/9uIx78gQEf/qIO37r5O/jctZ+HjJHB5XVhyD8088FyqFhdBK7fh1N/Pov+Ny5gpN+LkD+MYnX8qgSCIGAiMoFANICx8BgF03LoL0Nv4J2RdwAAb3vexpvDb0m3JauzKfACpq5RwMgY5FtiularRW9v77wEpqxWK9rb23Ho0KGE5B2WZWE2m5NmxhmNRrjdbjidzqTH5TgOTU1NFEybxGazZdQ+H567tMPKFRUV6OzshCAIaGhowAMPPICDBw/i1KlTCW19Ph+OHj2KH/3oR2hpacHOnTvxqU99KqNgmtvtjpvy2dzcDKvVCuBKRFIMrHV0dKCqqipphhshS80p32moZLHMgQef+xIAYIt6K26vuSOuXenKEhyxvYmnvnEIF94chCAIKbNLCCFkKs2NG1CzfQ0AIBrmEQlGceee21CbJMM1KvCQM3KUqcqxpmTNQneVTFKmKoM36MWQfwgnRo4DAArlhUkXromEohg6MYzRwTEpiJYqu4QQMncYRob61Q3ouO0x3LtZL62qKGNkuPfqZphv/S7qVzfkfeCpoFSFInUReF5AgAsAAlBeXZaQ4RpbPTJ2XimakgFFFtZ/vv2LK39vkOE/3/5F7HtGlrRunxAVEAlGEY1EpfIOAi8gz/8054w4zdNsNk8bvDGZTAnbxNhEqqQfq9VK8YtJHA6HFO9ZTDJelKC1tRWtra04dOgQbDYb2tra4PF44PVeyXphWRYajQa7du2C2WxGbW1txh3TaDQwm81xUz0n02q1M6YCErLUeINePHP6DwjxITBg8Pnr/gavnn8FcpkcESECJRO7suQ+fAZjl/wAgP4jFxAaD0FVrEL/0Qu4c89tuXwIhJBFZEP9Ogi8gLGhcXz8O00oZouStvvmTd9CMBLARf/FvP8AuNSNhcYQjoYBCJAxchTKCwEwEARB+t0Ex0IY9/jhvTCKCW4C4YkIqq9bBSHCI+QPo7yaVo0mZD6tKFqBf77xmylv31K5ddrb80lJZREKS1UAw4Bdl7w0kEKmwFXlG+NWlCQL7y9Db+Atz5WMNB483vG8jbeG38L1K66PGyf4KI9oJPYl8AIEXoCySAGBjwXU5Mq5m+7pdDrR2toKt9uNQ4cOSdleXV2xrHeLxQKHwwG32w2O49Dd3Y19+/ZJAS4x6NXT0wObzQadTpfymN3d3QCQMsYwldiupaVl2nYtLS1ob29P2G4wGGC1WpMG1VwuV1oZVtk+PyKHwwGn0wmWZdHb2wuj0RgXT5kczJp6+1w9n3a7HR6PB5WVlfB4POjt7UVzczOAK9lp7e3t0Gg0Ccd1u90wm82oq6sDy7LzNr03U4wwh5WEvV4vKioWT/aLz+dDRUUFvF5vVrXhCFko/aP9+I+jP8Cx4b/gY5qPo/W9RgiCEAumya4E07oefRFsTTnGL/kRDkQAAGXVpbjx89qk2SWELCVzdW6nMYIsRuPhcQz5BxERIqguXoNSVWlCMM03MAploQLhUATcWR8CviCqatUoqiikVT7JkjeX5/a5ONaQfxBvXHoDwUgAWyq3oo7dlFWfyOIUCATQ19eH2tpaFBbOTxbfnhcfwTsj78QFNWWMDFsrr0H7Leb4YFqYB8PEputGw7Ggmlwlh0zOzNsqnwzDwGAwxGWDqdVqGAwG7Nq1SwruNDc3o7KyMiFIxTAMurq64qZlpjrmvn37oNfPXKOwrq4OHo8HIyMjs3pMTqcT9fX1sFgscdloDocDQGb137N5fux2O9rb29Hb2yttU6vV6O3thUajkRaHFANYbrcbdXV1cLlcUnBruj6k83yKAbHJ/ero6IBWq5WeB4Zh0Nvbm5A45XQ60dzcHDetV+yj2WxGW1tbWs9huq+zTM7tc/pKWEzBNEIWo5qyGnzn1kfx6K3fxT2b7wUQO/FMrnkgLkbQ8oOP4fO/bIH+/30ENdvWoKBERcE0QghZ4kqUJbiqfKMUTAMQlzUoLkTA1lRgpaYKdbdchXXvXY2iikKo11dQMI2QBTThDcDx6Ct46bHX8IdfHYJzMHmtJUKyJWanTc0Q5AUebw2/iWOX/nJlm7gQQYECykIlCkpUUBYpIJMzUKgU87YYAcuy0peooaEBDocjLsDS2NiInp6epPune0wxs2ombrc7q0wocVbd1ODf1MBfOrJ5flpbW7F37964bS0tLXEZYGKQD4jNFmRZNqH+WzbPp9PpTKg1l05QE4gFCU0mU9z9ajSauGBfruTP0hyEkLRdW3UtqopWJL2N6/divXYNGCZ2BanqKjUtRkAIIcsIwzBSMG2qaDgKZfGVizAyuQyFFYW0EAEhORCeCMP3+jjWnN4A9lIVgpFArrtElqjJtdOmkjEy/Oc7T0o/CwKkVVuB2PeMTLYgCxE0NjbG/cyyLBoaGhK2ZXvMdGk0Gng8nozubyqj0Qin0ykFpziOQ1VV1ayONZvnx+l0guO4hKyv+vp6KfBmMBik7DWO46S+Jnvss30+dTodenp6UFdXB5PJBIfDAY1GM2Ng0e12w+12JzzOTO57PlFAjZAlhq2pwFnnhSuFQ2kxAkLIPAlGAvinV/4BT771cxy7dCzX3SFpkCvlCPvDcdtoIQJCckPgr0Qo6ko24eZ1t+SwN/NjIjKBC+MXMBLwIBQN5bo7y1Kq7DSRmKX2l6E3ACA21ZOPj57lciGCXAZNdDodOI4Dx3EztrXb7Um3T12cINViBPX19WAYJu4r1Qqhk830/IhBM4fDAbvdLn1VVlbGZajZ7XbU19dLCyzMNjMv1eNgWRZ9fX3Q6XSw2+1oampCXV3djM+tmNWWD8GzZDJelIAQsvDeGHodK4pWYm3p2hnbbmhcB+ev/4Knv/Es1mvX4KzzAi1GQAjJ2Nkj59H7qzdQuqoEW5s2oeaGxNU7XV4Xjg4dwdGhI+CCHK5fcX0OekoAYCw0iiJlMeTM9IGxIrYQntMjEAQBqhIVwv4wLURASI6UrS7FtR/ajLd+fwJ1GzSorcj99KVM+EcmEPKHIVfIUFxZlDQwPxGZwHh4DOPhMShkSqjkNK18oYnZadMtCCFmqT26sgOMTIZoOIpIKJadJi5KMJcLESwWJpMJVqsVnZ2d067IOXUq41STFycYHh5OGhyaXN9sLonTInU6Xcopkh0dHbBYLOjq6sp6GmWqx+F0OuOmv4qLSbS3tydd1EDMTBP7k05QMxeW36uCkEUmwkfwfef38EWHEY/1dCDCR1K2neACOGo7BrlShuFTI+j+z9cRGA3izj23Uf00QkhGuH4fLr57Ca6XTsM/PJG0jdt75Q3k1eotC9U1MkUoGsKAfwCnvacwEph+agrDMBAEYOSsD1y/D3xUQHl1GdVOIyQHGBmDM93nAACnXzuHOVwrbkGEAxGEJ8IIjAZTtglFr9wWW3WYLKSZstNEcVlqggAIsRIBfDgKIBZMm6/aaflMo9HAYrHAZDJNG9Cx2+3T1gMzGo0AYrXApk6ZnG86nS5pPTTgSt00k8kEi8USF0wTp3tyHBdXX2223G53Qp02i8UybRZeZWWlVCstWd28fLD8XhWELDIv9r+AoYkh8ODhD49DIUudWHr8eTf4aGw1ni076vCFzk/j3v/zYQqmEUIyFvBdqeVTuqokaZuP1H4UP77zp2hr3IP61fUL1TUyxUggtvoYDx7A9HNyAqNBqIqUKF1RjBUaNS1EQEgODfeNYOySHwAwNjSO4b7ZrSSYK0L0cpCGAWSK5B8rVxdXY33ZBqwqXg2lXJm0DZk/09VOm0rGyPCfbz8Zm94pi9ViVhQq5nUhgpkkm2453RTMdLKY0p3CKTIYDNi7dy927tyZNPhjtVpnrAOm1Wqh0WjgcDjSLsSfjnSfn3379klTOUVutzuu3eTvxcwwjuPg8Ximnf6ZyfM5NRPN7XajqalJ+lmr1UqBMzGjDYhNl526r8PhgNvtxvDwcFr3PV9oyichea5YUQx1gRqhaAj6q3elbOc+fAbvdJ2UiogWryhaqC4SQpagxs9uw/bm92BsaBwlK4qTtmEYBiuLV2Jl8coF7h0R+cN+jIXHoJKpEBV4lBekXt49OBZCYDSIiZEJCEJsuhkhZGGM9HsTAmZvPv0uAEBALDPt+R+9im0fip86X1WrhjpP6+CyNRXgozz4qBC3mvBkDMOgQF6AAnnBAveOiNlp6eIFHm953sQbQ2/gOvX1kClkKX+vc8npdMJisYDjOLS3t8Pj8aClpQXt7e3o6emB2+2GyWSC2WyWpiaK0wXFzLHJ+wNXsp9SHVMM2kxdfTOVtrY26PV6KajDsqy0sIDBYEirvpfZbE57ddG5fH40Gg30ej00Gg2MRiPq62MXQCsrK6XgXldXl7SfVqtFZWUlbDabFITT6XQwGo1ZP5/Nzc3o6OiQni+O49DW1ibdLgb+OI6Ly5bT6XRSfxobG+HxeKTMNavVKv0N5AIjLLbc4jnk8/lQUVEBr9eL8vLUb0AJyYVB/yB8IS8ef/0/cHzkODaxm/DFbV9GuaoCq4pXxbV1Hz6DrkdfRM22NVivXYMzPedx7o0BNO25DRrKTiPLzFyd22mMIPkszIfB81EMTQwhEA2gUF6IqsIqKORKKGWJWSDBsRB8A6NQFSuhLFYiNBZCOBCh6Z5k2ZnLc3smx3rm0RfQd/hswnYBAhgw0v9T1d60AXeaqA7uUhcIBNDX14fa2loUFs7N1FjTi4/gHc/bUsA2HQwYXFN5Lf7lA+0QeB5ypXxZTvUkS1O6r7NMzu3zmqH2wAMP4Ic//OF83gUhS9buZ+6Tvr9n8704eOIAHn7+IQDAbz/xVFzbI7ZjqNm2Bh/+xg4wDIP3fPwaPP2NZ3HEdowCaoQQsgSd9p2SvlcXqDESHMG58Vgtpk3s5oT2/pEJqIqVqFgbe2NYzBbBe94H/8gEBdQIWQC3f+VGyOQyuF4+HbddDKIlC6rV3XoVbn3g/QvaT7I0BCIBHB95N6NgGhAL8L478g6isjDkUIKP8hRQI2QaWQfUDh48mDR1keM4dHZ2UkCNkFl6uP5r+D+9jwEA7qjZgYMnDkjbp+L6vWj87A1SWjbDMFivXYPu/3x94TpMCFlWXuh/HmOhMWyp3ILacg3ksulXlyRza3VxNS76BwAApaoyjARHpO3JRMNRFJTFlwJQFivh9yRfcIIQMrcKSlTY+bVbULN9LV5+/M+IhhOLxIvBNLlShlsfeD+u3qFZkCl382Uk4IFcpkCRvIjqpy2wQkUhfnznT+ELeQEASpkSYT4MAFAwCsimjNmRUBQyeaxuWomyBAWKQkQjsem8hJDUsgqo7dmzB1arFQ0NDQnzhvN1WVNCFgsGDKqL12DAfwEPPvclAMAW9VbcXnNHQtvy6jK4XjqN6z+6FTK5DIIg4KzzAtg8rblBCMlvb/3xBHp//QYCvgDK15Sj8bM3JGS7Pu1+Cm9frs3yiw/9EhUFdL5ZSGWqMpzxnYYAAWd8p8EwDArlhShTlSVtLwhAwBtAMXslqBb2hyFXUiCUkIXCMAy26upQvXUF9n/pf1I0AvT/9yN5/x5u7NI4Ri+OQxAEFJSqULqiJC7blRd4eAIeCBCgkhVgQznNmFho6kI1ihSFAMNIwVoZI4NKnpiVHAlFIAiAQiWXgrgCL2ARx3MJWRBZZ6iJy6kms2fPnmwPT8iyJAgCfv3urzDgvwAGDO677gt45fzLkMvkiAgRKJn4q3xVmkqceM6NH7f8GlfvqMXYoB/9Ry/gzj1Uc4MQkhn34TN46T/+LP0cDUXR9eiLcTUZw3wYJ7jjAIDKwkoKpuVAOBqGN+gFIKCysAqlqlIADAQhRXFwQcAlN4exoXFUblQjGooi5A+jvDp5AI4QMn+Y6abQCTPcngeCYyF4TnOIhqJQFCogRHn4BkbjajIGo0FMRCbABTisL6NgWi7wAo+owAOCAKVMdSVQlmSckMllCPnD4MPR2GIEMhkEXoBcmd9/i4TkWlYBtcbGxmlv37t3bzaHJ2TZcnvdODfWDwC4tuo6fGLzJ/FXmz4RC6YlKTY9PjQOAOAjPE481wf1BhZ37rkNtVQ/jRCSoSO2Y2BrysH1+wAA2+69Dn2vnpFqMg76B8EFR7CmZA3Ojp6FglHgJHci6YIpZP6MhcewsngVGAZQF1SiqqgqZTBN4AUwMgYlK4oRGg8j4AtCrpTTggSE5Ejfq2fAMLHM0akYBug7fAbb7rlu4TuWJv/IBORKGYoqYkW91RtYjA2NSzUZw3wYEAQUyAtQrCxCiA8iGAlAJpMnfR9L5kdUiEIhi33cV8gUkMvkqS+6IBZU4/nYNE+5LBZMo/pphEwv6ww1n8+XcuUDm82G3bt3Z3sXhCw7mgoNzLd+F8/3P4drK2NvqBiGSchME129QwNViQpjQ2O493sfWciuEkKWGK7fi+s/sgVqPQs+KmDt9asQCYSlmoyZLJhC5g9bwEIpU2I0PCpN80z1IUkQBBSri6BQxYJoZStLFrKrhJApTr58SgqmrX3PanzgPi0O/8SJC3+5CEEAXC+fzuuAWjQcRenKEhSUFECAAJlcFleTUVw0hWEYXFW+ESPBEZwdi61wmmzRFDI/lDIlokIUvMBLdU5TjRMMw8QCaDwDmYICaYSkK6uAmkajgclkAsuySbPVLBYLBdQImQWGYXBN1bW4puratNpv2VmHLTvrICS71EkIIRlgaypwyT2C9/1/28EwTEJNxkwWTCHzh2EYlKpKL0/1nJ5MLkOxugjF6qIZ2xJC5tfo0DiG3SNgZAze99fbcMMnrsVvTh7Eu3e+i0HtEDb9+r245PJgbGgcpXka/JYr5YAAFJYXSNsm12TMdNEUMj8YhoGCSe/jPiNjaHEhQmYhq4Dazp07wXEcWJaFxWJJuN3r9WZzeELIDNyHY9OwuH4v2JoKbG++PqFwOCGEZGJ78/XoevRFPP2NZ7FeuwZnnRfiajLeXnMHnnL/Du+OvDPjgikk94JjIfhHJhANRyFXylGsLqJpnoTkEMMA6+vXouFT78Wqq1fgpcdfw6BjHJXhDTjSfASGf7offQfPAXlcDL5YXQTfwCi8531QFisR9ofjajKWqcrgDXIIRAM4O3oGAKZdNIXkFh/lwUd5CELs71Mmpww1QtKVdYZaT09Pytvvv//+bA5PCJmG+/AZdD36Imq2rUHjZ2/AWeeFhMLhhBCSKc2NG9C05zYcsR1D93++DramIq4mY0SIIMyHoamow51X3YXnzj6bcsEUklvBsRB8A6NQFStRUFaEsD+cUDicEJI+38Aojh58Uwoc8bLojPscPfgmfAOjWFFXhcLLr7utO+swNjSOscs1cJlwLHhRMFGI/tLT+PA/3TlPj2BuFJSqUF5dBv/IBPyeiYSajDzPIxQNQ8kowRaqMRryYdpFU0jO8FEe0TAPRsZAJmcg8AKiYR4AKKhGSBqyCqjt27cv6fZnn30WlZWVFFAjJE0D4xdw4IQdVYUr0D3wZ1xTdS3uu/4LkDOpU6+P2I5h7fWr8cGHbkKxugjv+fg1+OIvvogVtkoKqBFCsrJeuxY1N1RDJpdBrpSDkV35AKSUKTE8cQnekBf2Ezb8+M6fpFwwhcy9QCSA8fAYSlVlKJAXTNvWPzIB+eW6aYyMAdgieM/7pMLhhJD0+QZGceDh3+Mz1k9Ir58XrX/CnZumr13rGxjF2388CeBkwm0r6ipRfeMKFK0pwJB8EAIj4Cn371DL1ub9Qi/KIgUqisrAMEzcGAHELrxEhSh4RBGIBFBTtp6CaQsoykcgAJAz8hmfcz7Kg2EAuVImtY2EouCjPAXUCElDVgG17du3J91eX18Pj8eDAwcOYNu2bdncBSFL3sD4Bfzt81/FvqYn0H3xNZzgTuAEdwIu7iTab+1IuR/X78W6G6rxi88fQHl1KSY+48G58jPg+mmqNSEkOz2/fB1v/NfbAICPt9+JNdde+VAX5sPwhmLnmXA0hIMnYzXUSpQluHvjh2Y89t8+9yD0V7fghpXbAAB/PP0HAMC9m/Vz+RCWLF/IB1/Ii5HgCNaUrEWJMnWNpWg4img4iktuDxQFClSsLYsrHE4ISd/Rg2/imrs2xQWjr/3YZtzz1Kdm3Nf43/8r6fE0N23A//rTp4C7r2z3+UYWxUIv3vOjiAQjAAOsrKuKuy3CRyDGcRQzFMMncy8iRCEIAiKIoEBeMO1zLwiAEOUhhACZjJEuovFRqstMSDqyXuXz1KlTcDqd8Hg8cds5jkN3d3e2hydkyTtwwo67Nt6NUlUpDp9/Vdr+5vCb0+7H1lRg8PgwAGBgbAC+6IC0nRBCsiFMeiMtk8e/EQ9Hw7ij5oN46dyLuLXmNikQ9tM3n8CBE/YZA2Murwvm7nbp57s23o0vbfvKHPZ+6RIEAePhMQAAAxmKFNMvMiBXyjF2aRwllcWIhqOQyWVxhcMJIelzvXwG7/9cfDKBqiSWmTtwbAjlN5Un3a/mhjUJ24Zcw1CVxKZNTl7oZbLFvNCLXCZHuaoCET4C1QyZtGRu8QIvLVImY2RpBTIFXgAjg7TyrMALoPgnIenJKqB25MgR1NfXg2VZAEBlZSUAwOPxoK6uDjabLesOErLUvXzuJXzuuvsAAH/X8Ah6BrrhHOxF1+lncHTwCLatSp4JKhYOV5Uo0bf5HNYeXQNoAG3z9QvZfULIElSxrhw129aAj/BQlcRPDSxWFqNAUYC/2vQJfP66v5G26ze34DNP75oxoHbXxrtRx24CAGxbuQ3VJYkfNklyDMNgQ9lVGAuPgReikDHTT8cpqiiE78IoJrwBqEpV8J73xRUOJ4SkJzgWQmg8hPLViavqjofG4TnFATcl31dz81UJ297+4wnc9sUPAIgt9PJb13/jJHdCun0TuznvF3pRFCoSLriICuQFeT1ddSmTMTKo5CrwAg8mjZUtZHIZ+HBsUQI5I0MkFIXAC5ArabpnrtjtdrjdbrS1teW6KyQNWQXUrFYrXC4XamtrceTIEQBXpoH29fWB47isO0jIUjYWGsN4eBzVl5cRL5AX4OZ1t+Dmdbfg1fOvwOV1pQyoiYXDf/Pcf2HNiRqENwcBQCocTgghs3Xdh67GdR+6OuXtky8EiEpVsQ+a010IAIDqkjVpTQ0lycllclQUpJeJXFhegOprV8E/MoFIKAI+KtCCBITMgu/iaMrb/OExhMbCaR/rTz9z4v3/n1b6OSJEIAgC1pXWYDO7GW973pK25/NCL2UrU083J7klY2QzXnARyRUyMEWK2CqfPMAwsWDaQtZP4zgOJpMJDocDbrcbOp0OGo0GLMuC4zi43W5oNBqYzWYpkWe6/UQejwdGoxE6nW7a+3c6nbBYLGBZFlVVVRgejs0A2rt3b9z9ZfoYtFqtdJvH40FlZWXCY0jGYrGkFVCb3G+O48CyLPbu3QuTyQSLxZJWv0n2sgqoabVa1NbWAoit+Llnzx788Ic/BADU1tbi2Wefzb6HhCxhF/0D0vdHB4/gJ8d+jPuu/wK2rdqOMmXZ5VWRUltVX4nr1mzG3Rs/hFfOvYwXuv84310mhCxzUy8ETFaiLJn2QsDkY5zkTqBMVSZlq5GZ+cN+XJq4hBVFK1CsLE5rn4JSFQXQCJlnIX96ATXfwChC46G416QCCnzi+Kfw3NvPYajQh5K7SmG+7bu00AvJWJSPIiJEoGAUkMvSn9ovky9sAG0qlmVhsVhgtVphNBphsVjiAmNALJGnvr4eNptNClTNtJ/b7UZTUxN0Ol3KAJPJZILb7ca+ffviAl1OpxM7d+6E2WyeMSA3uS92ux3Nzc2w2WwJgTOj0Yja2lr09fWlDKqJAUTxa+rzIHI4HDCbzejq6orbt7W1FU6nc8b+krmT1Stn8pzsiooKdHd34/Tp09I2+mUSkp7zY+fw5Ns/R5+vD0++/XOc5E4gKkQxHh6fdr8/nv4DZXoQQhbU5AsBU6VzIeDo4BG8PnQUm9jNAIB/fOXv4eISV78jV4T5MIKRADyBYYT4IDyBYQQjAYT59LNiCCGzV1CSOihdrEycBprK0YNvJtRUY2QM3C+cwZq+DVjZvwbnx85BwWRd5posI7zAgxd4RC8vRhAVotK2xUQsH5WMwWCAyWTCzp07095Po9HAaDTCarXC4XAk3N7R0QG73Z40+KXVamE2m9HU1AS3253ZA0nBbDZLmWypdHZ2SmWzpssyM5lMMJvNcdtYlsW+ffvmpK8kfVkF1ARBwJ49e9DY2AgA2LNnD3Q6HZ577jkcPHiQFiUgZAbi6myPv/FDHB85Dt1VTTg+chwPP/8QhiaGpt33yIAT2y6vkkcIIQvFftwOAPjFWz/HSGAk4faZLgR8aduXcfO6W1CqKkUduwl3b/wQzK+1T7vPcnfadwpvXHoDQxNDKFeWIxAN4OzYWZz2ncp11whZFgpKY4X1Q+OhhNtKVCVQFaeXTeZ6+QxW1CV++C9mYwuMlIbK8PX3/wN4LK5AyFRnR8+gf/QsBv2Due7KshCKhqQvhmHAC7z081Tuw2dg++rv8CP9L2F78HdwHz6Tgx7PjsFgQGVl5bQBqanEDK+pQTG32500KDWZTqeDXq9Hc3Pz7Do8hRi0m7qY42S9vb3QarXQ6XSw2+0p27nd7qTHYVk2rYw6MneyuvzR2tqKffv2oa6uDgCg1+vhdruxc+dOMAwTl4JICElUqowvDO04Hf+aEQNuU0XDUXRa/gvv99+MsQ+EcMMnr523PhJClh+n7RgGj18CANzx4I0oLLuyStulidgHpBPccRRMWb1tNJy6zpBo6iIEdWwdBvwDM9ZeW85WFa1G/2g/BAgoVZZKMwRWJ5l2KwpPhOG9MAplkRKF5QXTZtgQQqZXUKqCqkSFwFhigAIAqq9fOeMx+o9eiC1skGRRkCbTbVAUyFHEFkFZuDiy03wXx2LF6xUylE6qpyYIAkLREITL/8jcCQciAABFgVwaB6LhKBBmEGHC4OX85cy0KPggoJQpIcgEMLJY25Mvn8ah774EMAAEwHOGQ9ejL+KOr96I2g9sgLJQIbWNRnjwER6MjIFCdWX6aDgYAQTE1Vnjo/yCTRnV6XTo6OiYNhA2WVdXF1iWRUtLS9x2cX+9fvqFlHbt2oXm5mY4HI6sA1VigMxoNCa9neM4Ka7S3NwMo9EIp9MpTXGdrKGhAUajEV1dXQnTQjMJOJLsZf2X39raitbWVunntrY28DyPaDSKHTt2ZHt4Qpa0UlUpSpQlCavwbFFvBQBsW5n8w+VB50GMFnJ4vrQL/+U9gJ+++QT+cOr3AICfvvkEDpxIfUWDEEJmMnj8Ek6/1o/Tr/Xjkjv+Cqj88lQklUyVUMdrPDye8kIAEDs/TZ3eKV5YGJhmKulyV6gohOpyPSVPIFYsuVBeiDJV6tU6w4EIBF5AaDwEPrq4s10IyQd1t2yAbyD+osHYYCwjN52Amu/iWMrbKq9iUV5dtmiCaUAsWy80HkJoIn7qOS/wkDGxAIycpq7OqSd2/RpP7Po1Ar6gtO3137yFn33ahj//6CiAWIkABgw6/+Z/8LNP2zA2dCVr/PCPemLfiHFOAQADvPBvf8JPPr0fI/1eqe3xZ114Ytevceixl+L60Pnl/8ETu34d997A9dJpLBQx4DTTNEwxA83tdqO3tzdhSmdPT09aCw6IwapsEoU4joPVaoXJZILNZksZmLNarVKATwwA7t+/P2lbceGCuro6NDU1oaOjQyq3laruGpkfWZ/lTp06BbPZjJ6eHpjNZuzYsQOHDh2C1+vFPffcMxd9JGRJu2nNzXht4DUUKQqxWX01Bv2DiApRAEiZrXHVyc244bUw+AiPklPFuGlLA6rXrcHrQ0fx+ev+ZiG7TwhZgoRJAZhTfz4bV/Pn0ds68OmnWvCpLZ9Oum+qCwEAcPDEAVSXrIlbiGDsclZbskUOSIyCUWBVcTXCfBBsQSWiQgQAA0EQ4urZioJjIXDnffB7JsDIGBRXFi18pwlZYrbdcx2e+udD+MDnrmSLnHz2NJ58/Ql8Gh8HEHvtdXW8iPd/bjtW1lXF7T81GLdUTD0DyWVy1FbUQhCERT91dTFhwEAlU8XGhBSJgRO+QOJGIZZVuFiIQbBkBfvtdrt0e1dXFziOS7rAgbj/dDXbpuI4LqN+Wq3WuICdy+WCXq+fNtjlcrmk28Wpm1arNWk2nkajgcvlgtlsRmdnp1QjTqPRJM1aA2I140wmEwwGA+rr68FxHLq7u9HU1ASDwZDQf6PRKLUVn4P9+/dj165dM65AupxkFVA7cuQIdu7ciZaWFhgMBukPbefOnejr68PBgwcpqEbIDJq3tOCNS6/D0vQjMEzsA9JP3vwx7tx4l9RmLDQGc3c7Pn/dfWDeVuFI5zHUbFuD6mtW4PxfBtH16IuQfSXJIEkIIbOw4+9uQeeX/wd+zwRO/bkfN7c2xgVubll3K0aC8fXTBsYvAEh9IQAAPnfdfQkLqRwdOooSZQlN95yGTCZDHVsX9zuYLpjmGxhFiboI7LpyBLwB+D0TUKgUtNonIVkory6Dru1W/OlnTqzaVAXfxTGoSpV4+fRzUpvgWBBDJz0IjiZODS2vLkN59fQLGAiCgJHgCHxBHzZWbJzrhzCnKq9iMV0chmEYyJH+SpNkZn+z/1MAYlM+RTd88lq85+PXgJEBCsWVj/Z//TM9GCZ+uqa6pgKeM1x8wI0B1Osr8NFvNaGg5EotwKt31GHTbbXSFFBRyw8+Jk35FNXdetUcPcKZifGGZAGjyQErg8EgrQyabFXNysrKaWuZicQ26WSzTWYwGBL2cTqdqK+vh8ViSQhgidlmk4lTTVNN+9RoNLBYLFK2msPhgMlkQlNTE1wuV9I+mUymhMUO1Go1Ghoa4u5DzKIzm81xj0Ov1ydd4GE5y2rKp9VqhcfjweOPPx437RMAamtrMTw8nFXnCFkOqkvWYM/7vo6fvfUTvHLuZRw8eQAVBWzch86x8ChOcicwGhrFEVssmPbhb+xA/aduQMPe69H/4RN4bugQAODR174jTf8khJDZGL04Br9nAgAwfsmP4b744Nm9m/V49dwrcdv+cOr3+OK2L0s/j4XGElbwrKuowyvnXo5rc/C4HV/e9uB8PIwlZWrwLFkwDQD8IxNQFStRsbYcxWwRKq9SQ1WshH9kYiG6SciStrKuCh/4nBaam6/Ctnuuw7Uf3Rx3e3l1Ge77ZQtqtq1J2Pfauzbj05ZPJD1uYDSIs0fP45F/M+Erv/wKvvPav8xH9+eUTC6DXBH7IgtDWaiI1TmbdP6XK+VQFiqgUMXnyaiKlHE10QCg/tPvlaZ5ApBqqTV+5gaoipVxmWpyhezyceODosqCWB8m10xbqPppAKRAUTrTGsWEn/b2xIWPtFotOI6bMfNMnEbZ1NQEAKivrwfDMHFfYpuZaLVa6PX6pDXU7HY7XC4XTCaT9NXb2wsg+WqfU6e8ajQaGAwG9Pb2SsG1qRwOR9LAHMdxCcdzOp3QaDRJA5GZZPYtB1llqCX7hUyW6s0eISReHbspbgrUVNUla/Crj3QCAHr730HjZ2+QXl9rSteguXoXuv/zdXyhM/kULEIISWWk35sQMHvz6Xfjfn5lXzeu+/CWuG0PXPUV/PTNJ7CZvRoD/gGUqcpTXggQbVu1HUcHj+Cnbz4BABgYH8AXt32ZstPmUDQcRUFZ/BRPZbFSCpASQvLPxXeH8IdvPY/NeC+EbQxclW8hHA1DKU9v9dCFFhwLYWx4HKGxMFSlSpRWlVAG7CKguXEDmvbcht5fvQHveR/YdRWo/9R7UHvjhtgiBNH8n/rpcDhmXEhgqmQBr71798Jut6OzszMhW2yy/fv3Q6PRSBlbYpBrtsRA4NSsM5fLlTJw1tnZmXCbxWJJORVUDBZO1dXVhYaGhrht4jTZqXXdurq64rZZrVYp6y7T53+pyyqg5vV6436eOv+6p6cHu3fvzuYuCFnSBEHAK+dfRm2FBmtK1kDGzHyFh62pwFnnhVh69+UpomedF8DWVCxAjwkhS033k0fRd/jstG0G3hrCwFtD8KwawmDNeRT6i6FdXY/PP5y6ZuPkCwGTbVu1nQJoGYjwEQSjQRTIC6CQzfy2Ta6UI+wPA+yVoFrYH4ZcSVOvCMlXxZNer+uwHls1dQjxobwMqAXHQhg6OYxwIAxFgQKBCwFMjASwclMVCkpV8IV8iPARKBg5SlVlab23JdnhBR6CIEDGyGZMaNHcuAEb6tcCiJ8OKvAC8j0Xxm63S4sMZGJy9hXHcWBZFlqtFm1tbVJNsVT353Q6sw6iTSYG9yYH09xut1SnbCqj0QiHw5Gwyqjdbk+50inHcUkTnxwOR9w+DocDFosl6aINYjab1WpNeju5Iqsz3Pbt29HY2Ijf/OY3OHXqFEZGRnDq1CkcPHgQmzdvxv333z9X/SRkSfIEhtHR/SgecBjwnT9/K619qq9dhf6jF/DrB36L7v88iqe/8Sz6j16Atvn6ee4tIWQpuv0rN6LulvTqn3iqB+G64S28eWMPVn+CgvgLYSIygQvj53HK14eRwMiM7YvVRfBeGMX5vwxg+MwIuHNehPxhFKtpYQJC8tXwGQ6FFQVgZAyqB2qwc/zuaVdMzqWx4XGEJkJQFSshk8tQWFaAcCCMseHYapJjoVF4AsMYnBhcVMXuF7MoH0WYDyMYDYIXZl4IQiaXIRqKIDQRRiQUQSQUhcALCzp1M5npaprZ7XZpGuTU4I4YMEu28qdOp4Pb7ZYytqxWq3Sb2WyGwWBAU1NTQkaXWI+sq6trxll56bLb7QlBLbEfUzPHRGI2mM1mi9vudrthNBoT+m2326HT6VIuxOB2u2G1WmG321FZWZlyAQO32x33/IhTXkmirDLUdu7cCZPJhC984Qtx2Wosy8JqtWLbtm3Z9o+QJe3kpNpC68s2pLVP+PLy5L4Lo3j9N29BvYHFnXtuQ+2N6e1PCCGTFZSosPNrt6Bm+1q8/PifEQ2nfjMeLL0ybXA1u3ohurfsBaNB6fsCecGM7QtKVVAVqzDu8cM/MoHKjWqUV5fRdCxC8pT78Bm88P3DUj2r0Ytj6Hr0RTTtuQ2aPHxvFxqLZaaND8fGA4VKDmWxEqGx2PvTCB8BADCQQS6jzNiFIExaTZVJWHc1EcMwYGQy8FEeQpSBXCWDXCnLWUCN4ziYTCap7pfRaJSCWGKQTKPRJATTxPpoHMfBYDDAZrOhq6srLmBls9mkmmR1dXUJUxvNZjOcTifa29tRVXVlZd7h4eGMMrPEvtjtdgBAa2urFKiaXKNs8lRKp9OJ1tZWOJ1OOBwO2Gy2uOCd2C/gSiCwubkZOp0OBoMBFosFHR0dUn85jkNdXV3SqaMOhwMsy6a1OqfYVuy/OI1UvG3qc7jcMcIcXToQ0yE1Gg127tw5F4ecdz6fDxUVFfB6vSgvL891d8gydG7sHF4+9xJc3EnctfFu1K9OfnVist888gcMHr8ERsbgvl/vgrIgq7g4IUvOXJ3bl+MYwfV7sf9L/5P8RgYIfWUEXaN/gHZVPb7W0IZS1fQr1pHs+cN+TET8CESDqC6unvEDqsALuOSOXeVXFCqgpnIAhMSZy3P7XBzL9tXfwXOaS1h5seoqNfTf/0hW/ZsPw6dH4LswikjwclaTIhaMKV9Thqqr1AhGAgjzEfDgUa5aHmPnXAgEAujr60NtbS0KCwsz2jfKR6XMtHSmCfNRXrp4JlMwkCso8LnUmUwmOJ1OdHV1pdXW7XYnZMUBV2qpLVbpvs4yObfP2SdxrVabkA556tQpbNy4ca7ugpAlZ13pOuza8qmM9vl4+53wnvPBe2GUgmmEkDnFTHd1WgDeGn4LUAFnR8/m7XSkpaZYWYxiZXHa7RkZgxWaSkRCUYCmWxGS97znfPHBNAAQAO6cN2n7XCutKsHESADB8RAUKgX4SBSFZQUorYqNCQWKQsycS0vmklwmhxzpB8VkchkYGXO5blqeF04jc8LpdKY9bTNVW6PRmLJu23I2r5/GzWYzfvjDH87nXRCy7MgVMlRexaLyKjbXXSGELDF9r54Bw6SIwzBAwbslwHuAoYlB9Hnd0LB1C95HMjNGxkBZSBdcCFkMKtaVJ2SoCRAwWp6fAbWCUhVWbqpCkbqQVvlcxBiGASOnYNpSx3EcrFYrHA4HNBoN3G530pppydqK00w5jpNWO6XFCRJlNeXT5/PBZDKhp6cn4TZxrnA0Gs2qg/NpOU7nIfnFH/ZnlHlACJkZTfmcPfvfPoVhd6zwffHVBVj9STUuHhyB/0QQAgSc2XISx27qAQMGt9XcjvdVvz9u/9qKWtSUrc9F1wkhJC35NuXTffgMuh59MWF7z46X8L0HHkOpqiyrPi40f9iPSxOXsKJoBb3HzUA2Uz4JIenJuymfu3fvBgC0tLQkRCtHRkbiVtEghMR7Y+gN/NMrf49/veP/oo6yPAghOTY6NI5h9wgYGYPx2z146qo/AIMAbgbWV23Cur4NOHXNcQCx7IkX+p/HC/3Pxx3jprU3Y8/7vr7wnV+iInwEckZOU3IIWcI0N25A057b0P3k6+D6Y1lpvsoRyK7lwQW9iyagFubDCEWCuOi/iDAfhicwDDkjg0wmh1I2c10vMju8wEPG5HZ1TkKWs6wCao2NjXjkkUdS3k5vAAlJ7cm3fg4ePP7xla/jn278JrZWbp1xn7f+cBzKIiVW1KrBrq+g1xghBL6BURw9+CbKq2MfulQlKlx71+a0933rjydQWFaA0YtjqNKocdsD70dJbRHCR8fxyplXIOMZnL3mJM5ec3LaY9267jZ8cduXs3485IqL/osIRCagkquwrrRmxg9NgiDAPzIBhUoBRaECcgV9yCJkMdDcuAG1H1gP+1efQmFlAd53wzY80vSVXHcrLQIvAAxw2ncKvqAP3pAX5coyBKNBBKIBAMAmNr0xiWRGEASEoiEwAGSMPK0FCQReAM8LYJhYeQD6LEFIdrIKqM00h3a6YBshy5kgCHB5XQCAsfBYWqsgCYKA7idfR2A0iGJ1If76p/r57iYhJM/5BkZx4OHf4zPWT0j1a/70MyeOHnwT2+65btp9+49ewNt/PIEm020AgOBYCL/7JwdWXb0Cw6dGsPL/1OLD/hqc3erCX27shpBQNRtgwEAhU+CLN3wJOzbo6I35HAtGAhAgIMqnl4EQDUXh90wAAArKClC+mlZhJWSxYBgGzf/vo7nuRtomvAGMD/sh8ALKq0uxurgaXJADABQpS8ALsbI/q4urc9jLpU0clzOp3yTwAvhIbIVPuUIGRkHjNiHZyOrSpUajwdGjR1Pevnfv3mwOT8iS5fa6EeZD0s8TYf+M+7z9zEkERoMAgHAgCvfhM/PWP0LI4nD04Ju45q5NccWgt997Pf78syPT7hccC6Gr4yXc9qUPSNuGTg7DNzAGAFAVKRHyhwEA7yv/ABQppusoZEp8/4M/wM6rmiiYNseiQhTFyhKoZCoUKNJbM88/MgHfxTFw53wYuzSO4Fho5p0IIWQWGCa2SiQA8FEBZaoylCvLUKwownh4DDJGhkJ5IcoWyZTVxUrGyMAgvUwzPsojEoogGo4iGokii1LqhJDLsspQ27lzJ/bu3YvW1lY0NDTEZaxxHAeHw4H29vaMj+twOOB0OgEALpcLdXV1aGtri2vjdDphsVhQV1eH4eFh1NXVwWAwZPNwCJkX/aNn0efti9v22sCfwYCRriz9l+s30xb3dh8+g5f+48+o0qhRoi7C2CU/uh59EU17boPmxg0L80AIIXnH9fIZvP9z2+O2icG1/qMXULNtTdL9jhw4hpWbKuMCcTXb1uC+X7YAAArLC1CxtgyF5YUoXVMcdwFgsjAfglwmn4uHQqaQM3KUq8ox4B/AClXFjO2DYyFMeAMoX1UCRi4DI4tlMJZXl9Hqe4SQOSdTMFCo5GDkMsjkDARBQLGyFMXKEpSpyjEa8gGIbacLLvNDEAQIggCFTDHjWMxHeUTDPGRyGWSKWBCUj/JgZAxkcioPQMhsZRVQu//++9HZ2YmGhgYMDw9jeHhYuo3jOHg8noyP6XQ6wXFcXACtrq4OLpcLFotFatPc3Ize3l4piGc0GtHR0ZEQeCMk137x1s9x+MKr07aZqbj3Edsx1Gxbgw9/Y0fsiqAg4OlvPIsjtmMUUCNkmQqOhRAaDyWd1qcqUeGS25MyoNb36hlcc7nOWv/RCygoU2FlXZV0u7JIiU/98K8AAAeO28C8FbsAIF4ImPz/q+dfwb2baQr6XArzYfB8FEMTQ+gf7YeckaOaWT1tcW//yAQKywpQsfZKCQHveR/8IxMUUCNkkekeeA3vXHwXF8MX8HD91/Ky6LyqWAXVhvhzy7rSdVLwrKKggoJp84QXYlM2I0IEIT4EMAAjxJ7nVH8rYvBMoboSeIuEouCjPAXUCMlC1q8ej8eDZ555Bp2dnXFfzzzzDFpbWzM+nsVigclkitum0+niVgxtbm6GXq+Py4gzGo0J+xGSDx7UPoRb1t2a0T63rrsNX9n+Velnrt+L9do10psShmGwXrtGWg2KELL8+C6OprytsEwlTRFPuu/lqZ1v/fEEVm6KBdJ+908ODLmGE9q+dO4lKZv2+hXvwffu+D6uX3E9gFj9lpf7X5z1YyDJ9XndeMvzNi6Mn0exsggXxs/jLc/b6PO6U+4TDUehLI4PtimLlYiGo/PdXULIHHrtyaN47R+Owfe/I3jpzIt48ezzue5S2qYGzyiYNj9C0RACkQDC0TBkjAzhaBiBSAChaOpp/oIQW4RgMkbGgGZ9EpKdrAJqTU1N094+mxpqTU1N0Ol0KW/nOA5utxt1dXVx27VaLYDYdFFC8kmJsgSPNJjw4PaHoJQpwSD5mwsGDJQyJb66/SF8raENJcoS6Ta2pgJnnRekWgeCIOCs8wLYmpmnARFClqfQePI31r6BWCCu/+gFXHvXZhSUxrLTtt1zHX73j4fi2g75B+H2uiBjZPjcdffhWzd/G3XsJnzr5u/gc9d+HjJGBpfXhSH/0Lw/nuVkTclalCiLwQsCihWx/0uUxVhTsjblPnKlHOHLde9EYX8YciVNySVkMRk6dwmKIRUUESXKRlj86p1f4SR3AoP+wVx3jeQJpUwpZaJN/j9VBjMAMAykmnci4fJqn4SQ2ctqyicA+Hw+lJcnX6HQZrNh9+7dGR1Pr9dDr4+fOtLZ2Qmz2QwA0jTSysrKpPs7nc5pA3KE5ALDMNBd1YStldfgq899JWk9olhx739DTVlNwm3bm69H16Mv4nf/6EDNtmqce+Mizr0+gDv33LYQ3SeE5KGCktTT+AKjMxejX1EXP47WbFuD0HgIb/3xBK69PB2UYWSoX92AT2/9DK5Wb8FjPR3oH+2HQibHY7d/D9eveA9+/e6vKAthjpWpytA/yuPSxCV4AyNQygtQULpu2uLexeoieM/7wPV7oSxWIhKIIOQPo7yaCoITspj8MfI0rmG2YVTNQRFW4IL/PB5+/iEAwG8/8VRuOzeDCB/BwPgAeIFHkaIIK4tX5rpLS5JcJkdEiCAqRBGNRsELPJQy5bR11GRyGaLh2KIEjCy2oITAA3IlTfckJBtZBdQ0Gg1MJhNYlkVjY2PC7RaLJeOAmshut6O7uxtOpxM2m00Kkmk0GgBIqM/GcRyA2CIGqQSDQQSDV6bA+Hy+WfWNkNmSMbJZFffW3LgB1deuxPm/XPz/2Tvv+DjuMv+/Z2abtmnVbMndkp3mJLYlm/RqOQnkArnEdsJxd1wObFHud0AuWM4Bd9zB4cjkegHJARIgkFiKjwsQCFIK6cG2XBI7xZbcLVlltVpt3ym/P0a71qrLkiyX7zsvveKd+c7ssyPtfGc+83mehxPvnCRnTja3bbiR+aJ+mkAwYZxrc4TdbXZ+HMyJlggnsA0huKW2G0po6WjqBBay/+WDHN15guXBmyhaYAr9h4OHORw8lH4KfnHuJfz9Nf8w3o8i6IdhGCSNJDn2HDw2NzbFjoExbD0iu9uGpup0Hg6AYZA7N0c0JBAIJpgzMU/cveZO/uPif0e3ZKZrP1j20IS/13gJd0bSaeXeQg8SEjEtCoBFF+7YySKVsWKRLUhIaIY24hwhKzKGAclYEgyQFQmL3SLqpwkE42Rc36AVK1ZQXV2dFs76/6Q6dZ4Oq1atoqqqKl0brbn5VN2Q9evXU19fnzF+NKmeGzduJDs7O/0ze/bs045PIDgd3jzx+rApn2+ceH3Ibd35p1JA7/j6LUJMEwgmmHNtjrC7bdhcNmKhwUX6WYsHb0iQ2m6olNCU0Na2v8MU1RpPEO6MAOC2upElmSxLFkktOej2gvEjSRLzPPMochfitrmZ4Z7BXM/cEZ2ANqcV73Q3vlnZ5MzOFmKaQDDBnIl54pYFt1KcX5yxbIFvITfNunnC32u8JMIJ4qEE8d75JJV+aF7rCufyZCFJEjbZhk22pVM9bYpt0DlCjavpf8uyhGJRUKzmz2BiWt/x5xINDQ3k5ORkaAYTTXNzMxUVFaxevTrdEHHTpk2Aae5J/buuro7Vq1cjSRI5OTlUVFSkzT99SY0pKSlJbys49xi3Q2379u1Drv/c5z43nt0DprC2bds2ysrKOHjwID6fj6qqKlavXk1dXR2rVq2ioaEh3aCgf221vjz88MM8+OCD6dfBYPCsv2ESnF/0Le59ke8iHrj8Mzz53k94t/PddHHvobrl5czxMXNxIRhkdOgRCAQTw7k4R5RcPyddEy1F6vVQHT5T27Xvz2xAEO8V5mYuLgQgK9uRXhcLmo6Mf7j2W2YtSJHiOem4bC7mWuahyMqoO/xZHRazJo5wHAgEk8KZmCdUwxQ0FvgWctvc2/nd4efTy63S0DWypgLzXKOBQbpbZHF2yVnZlfR8Q5IkLLIFS+/t/GDz8nu/28+urfu461vluAvMB/PpxgSDjA+1h/nlNxpYcs9lXHrbwskLfhLIzc2luLh4yLJQ42XTpk08/fTTVFVVZZSXCgQCVFZW0tjYmK7pniphVVZWRiAQoLq6etB91tbWsnLlSmprazOaLQrOLcYlqG3evHnY9RUVFePZfZqVK1eyadMmampqWL9+PWD+ATY3N1NXV0dpaWk6FTT1hzwYdrsdu90+ITEJBGOlb3HvP7vs0/zxgnuQJZlvX7+R/93/DD9578fp4t6D1ZwoXX05pasvn4LIBYILg3NxjlhyzyJ+/fcvcPWnT819+57fzw1fuCr9Oh5KUL/pFa769FIKSsyOnlf9eSlb/+Y5gq09aUfa2z9upPjaOekxl6xcwPxr55DldaSdTjZFOJ7OFIqkoChje3gi6qUJBJPLmZgnrLKVqhu/i4LCifBxImoEZYSC81OFZ9opkSbldhJi2pljuIdbalxl19Z9BFt6+OXX67nr2ytxF7iGfCgfag/zy6/XE2wNsWvrPhbeNB+Lfdzl1s8YpaWl7NixY1L2XVNTw8aNG9Pmnr6kzD4lJSUDdIiKigoqKioyxLa+BAIBVq5cKcS0c5xxfUuWLl06rvWDkZOTw8MPP5wWzuBUA4L+9dGKi4vTQlrKpSYaEgjOVvoX904hSzL3XrRaFPcWCARjxlvooXz9Dbz1RCPTFuQRPBnC4bGnmwoAxENx2g/4ifdpVGB327jnnz/G2z9uTNda8xZ6WPKFRekxre+3s+OpPXQfD5I900vZ/VdSLFLNBQKBYNKJdcR59/n3efHNl2mZc4To0iAfL7n7rLtGVOMaoc4wiVASm9uKO88lUs3PEix2C3d9qzwtkvUV1frTV0zzFrq561vl55SYNpmk0jyrq6uHFb4qKysHaBXr1q1LbzuYS62mpoZ169ZNdMiCM8xZ9U1J5RanRLIUqVzosrIywFR7/X4/tbW16THV1dU8/PDDZyZQgeA0yM/KH7Z4tyjuLRAIToeCkry0q2wwvIUeHvjZmgHL7W4bN37h6kG3aX7zCPWPvGKWwDHAfzhA/SOvsHLDjUJUEwgEgkkmEUmyp+598ikkYY2z85ImjvYcZY737Dn/xkMJ2g90kowlsdgtxFpiRLtiFCzIE6LaWYK7wMVd3145rKg2QEwbQnSbTBobG1m7di3Nzc288MIL6Xv/VM306upqGhoaaG5uJhAIsG3bNjZv3pwWuFKi1/bt29PNDIfa57Zt2wCoqqoaVWypcWvWDLyO6suaNWvYuHHjgOXr1q2jpqZmUEGtqalpVO608R6fFA0NDTQ2NuLz+dixYwcVFRUZzrmampr0v/uvn6jjeT4iGak2IWcJlZWVVFRUZIhqK1euxO/3p22cFRUVlJWVpRXdhoYGqqqqBjQqGIlgMEh2djbd3d14vd6J+xACwSg41nOM5u4mumJ+rpt5A/lZ+VMdkkBwXjBR5/YLeY6o/dKv8B8OQN8rBAk8s114v6wQSoYom76MhTkXTVWIFwRRNUpST6LpGjmOnKkORyA4L5jIc/tkzRO6pvOjTz6NGtewzJa5feMNzHTPOqscap2Huwi29GQ0zQp1hLHlW3HOcqAbOj67b8gO9oJMYrEYBw8eZP78+TgcjpE3wOz2qRs65mQtDXmshxLNzgYxrS+SJLFu3TqqqqrSYlBOTg7r1q3jvvvuS4s7q1evJjc3d4BIJUkS9fX1GRlrQ+1z8+bNrFo1eN3qvpSUlOD3++nq6jqtz9TY2EhZWRnV1dUZbrRUQ8WxZNeN5/jU1dWxcePGjLTYnJwcduzYQXFxMTU1NTQ1NaWFsebmZkpKSmhqasrQZcZ7PKea0X7PxnJuP+uS3KuqqmhoaEh391y9evWAnOiqqip27NhBZWUllZWV1NfXj1lMEwimmt8fe4lHt2/iB+8+xuHgoRHHN255h+e/83t+8+2X0sXDBQKBYDLoPh7MFNMADAidCFPzTjU/e/9J3vO/NyWxXUh0Rjtpi5ykM9bRe9M0PMGTIbpbegi1h89AdAKBYLKQFZk7vn4Lf7L5bv7yPz/JLM/ss0pMA0iETGealtSIh+LEQwksdgvdwSCdsQ664v50gwXB5JHUkyR1Fc3QhhyTcqp5prsJtoZ49mv1tOw9eVaJaWDWI0v9pFi2bBkNDQ0ZTqrly5cP2hhxMLfXUPtMOatGorm5eVyNDkpLSyktLR0g/vUX/kbDeI7P2rVrB2TzrVmzJsNZlhL5wMwY9Pl8NDY2jiqG0R7P85GzKuUzxUi5xD6fb8huGQLBuUKO49TJuTPaOcxIk5MfdnBk23EAOpo7mXnl0B38BAKBYDxkz/QOdKgBjsJTxbhDidCZDeoCxCIr0HuPpOlauuj3UCTCCQzdQBGdoAWCc56ZVxZOdQjDYnNbiZ6IEu2OgQGKVUa2yNhzrBi9Jy5N10CcjiYNSZKQkDAwGCnpzF3g4o++Xc6vvl5Pz8kQz/6taUY5W8S0FMuXL8947fP5BpSDGmsR/8H2OVqKi4vx+/1jer/+9G9OEAgEyMsbulTHcJzO8WlsbCQQCAxojJByzoGpv6Q0mEAgkE7pHOyzj+d4no+cdQ41geBCYVHeIh5Y9Bn+puyrXFmweOQN9FMT5cG3jk5iZAKB4EKn7P4rBzrUgKX3LeIrpX/DN67+e26dc+uZD+wCw2Pzkp9VQKGzEFke/SXb2eVjEQgE5yPuPBe2LBuJcIJEJEGoPYxiVSgozGe6s5AZrpnYlXOrc/a5iCIrWGQLFnlkn4w738WNX7wmY9ktX77urBHThmIqBZvy8nICgUC61vtw1NXVDbo8JVSlxKuhmhGUlZWZImmfn/4OscEY6fik3GoNDQ3U1dWlf3JzczMcanV1dZSVlVFZWQkwLmfehcRZ6VATCC4E5nrnMdc7b9Tjb/nKddT+9a+I+KMcevsY161dftbZ/wUCwflB8TVzWLnhRt5+Yic9J3twF7hZ/qnFLLxxPnDZVId3weCyju0mJ3euD8MAMTUIBOcHzW8eYcdTewgcD2IpkHHfbmP13fdMdViA2dimYEEesk0mHkzgzHWQXeTF4REi2plkNEJainBHmFf++62MZS/92+tnlUPtbKOyspKamhq2bNkybBZdytE1FH2bE3R2dg4qgvUtcTWRpBxs5eXlA9xsKTZt2kR1dTX19fVDjhEMjnCoCQTnCKH2MBF/FIBwR4TOg6dXHFMgEAhGQ/E1c/jk9z/Buv/9U/6k5m4W3jR/qkMSjICsyCgWecTUUIFAcPaT6rbsPxRAT+rETyTx/yjCgTcOTXVoaexuGzMum878q2cz/aICIaadxYTaw/zqGw30nDRrpn3ikdvxFrrT3T9F7c3BKS4uprq6msrKymFdanV1dcMW5a+oqADMhgH9UyYnm/Ly8kHrocGpummVlZVUV1dniGmpdM9AIJBRX02QyaRecX3+85+fzN0LBOctXce6OfDqoYyf1zdnFnt8ffO2AWO6jnVPUcQCgUAgmErioQRdR7vpaPbTdbRbNK8RCM5xdjy1J+N1qlbWWz+bHBeL4PylbzdPz3QXd/zdLeSX5HDnP64460W1wdIth0vBHE1q5mhTOFOsW7eOhx9+mBUrVgwqStXU1IzYYKC0tJTi4mIaGhomtBvmaI/P5s2b06mcKZqbmzPG9f13c3MzxcXFBAIB/H7/sOmfYz2e5xvjTvncunXroF0dAoEAW7Zs4Xvf+95430IgOK+JqlG6490Uuk4Vn932010cfHP4Ommt+9pp3deesWz+tXO4rfLGSYlTIBAIwDxnhRIh4lqcWZ5ZUx3OBYGmaxgYQ6b2xEMJgq092JxW7J4skpEkwdYevIUe7G7bGY5WIBBMBN3HgwOWSUjEWuNTEM3Y0AwNXddBAqtsnepwznv6NiToXw6mv5j2sW+uwDPNjaEbZGVncec/ruDXf/dCWlSbivTPxsZGqqurCQQCbNy4Eb/fz5o1a9i4cSPbt2+nubmZyspKqqqq0qmJzc3NVFRUpJ1jfbeHU86yofaZqis22kaH69evZ9WqVemaYz6fL91YYN26daOq81ZVVXVa3TDHe3yKi4tZtWoVxcXFVFRUUFZWBpg10lLiXn19fXq70tJScnNzqa2tTYtw5eXlVFRUTNjxPJ+QjJFaggzDhg0bqKmpYdmyZQP+iAKBADt27KCzc+TuhVNFMBgkOzub7u5uvF7vVIcjuABZ/8pDvO9/D4tk4ZmP/yI9CcbDCV79n7dpeu3wqPdVcsNcbvj8Vdhd4uZJcGEzUed2MUecIhlXafugA8Mw+LvWDXREO8ix5/DER3861aGd12iGxqHugxgYOC1OZrhnDjqu62g3siLh8JrpVpIsEw1E0TWDnNnZZzJkgeCsZyLP7ZM5T9R+6Vf4DwUyF0qQNy+HVf9254S+10SgazrJqIqu6ByPmw+F3VZPxgNjwdDEYjEOHjzI/PnzcTgco95O0zWSehJgQHOCgWLarXinewBTeNNUHTCIdcfT4862rp8CwUQy2u/ZWM7t43aoDddGdsOGDePdvUBwXuO0OAFQDZWeRBCv3bzxsbtsrHjoemYtncFr338bLakPuQ/FKnPD56/ioluLRZMCgUAw4cRDcX786WfQVZ3plxTgut1NR7SDUDI01aGd9yiSgtmz00DVtSHHaUkNuyeLUHsEXdORZImsHAfRrtgZi1UgEEwsZfdfSf0jr6ROAen/l913xRRHNpB4OEGwpQcAe7YNek1pujH0eUswMfS99u/rk1HjKr/8RkNaJLvj727BU+BO31NIsoQkS+iagbvAxV3fXpkW1X75jQZW//udWOyif6FAMBLjqqE2UkG9hx9+eDy7FwjOexb4FnB53hXcOOsmtH4XHZIkcUl5ifkUciidTIJV/3YnF68oEWKaQCCYFOxuO55p5pPq9gOdLPYt4doZ13HL7FsHnLcEE0+WxYHT4iTLkjXkGMWqkIwksWaZNz+GbhDrjqFYlTMVpkAgmGBS3Zbz5uagWGXy5uZw24YbmX/NnKkObQB9hRctruO0OHFb3cOetwQTg4SELEnIkpxxL2CxW1hyz2V4izzc9e2VeApcGV2gDcPA0PX065So5i3ysOSey4SYJhCMknF/U4LB4JA2uNraWj772c+O9y0EgvOWP73sz0ccIymy+WRyMIze9QKBQDCJLLhxHmF/lKLLplG8aI640D6DDJXm2RdnThbBVtMdIikSGKAlddz57skOTyAQTCLF18yhuI+AltAS7Di5nfZIO3fM/+gURpaJYpGxuWwoVhlrlpUcl0g1P1NIkoRNGby76qW3LWThTfOx2C3omm660wwDSTZVNEM3M11SuAtcwpkmEIyRcX1biouLqaysxOfzDepWq66uFoKaQDACu9p2UrOnmnVXVrBk2tIB6w++cQRJgsGqHUoSHHzzCEvuWXQGIhUIBBcqyz65OOP1rrad/OjdH/DA5Z8Z9LwlmFgiyQitkVYKnYU4rc4B6+1uG95CD5GuKFpSQ7EqeKa5RUMCgeA8Qdd0Dm8/zpafP0NL3jGOL2nm1jkrsClnz3c8u8iT8TqSjNAR7SA/K3/Q85ZgYtF0DVVXscgWFPmUOzkljsm9D+B1TU871RSLnF7ef7xAIBgd4/rGrFixgkAggM/nG7SjQ3d393h2LxCc17RF2ggmunli7+McCx3lib2P47a58dqymeaclh534LVDaTFtxhXTufqBUt78USMt75zEMKDptcNCUBMIBGeE1Hnrp+/9mIPBg/z0vR8Pet4STAxJPYmua7RH2znWcwxFUiiUpiPLyoDOeXa3TQhoAsF5Srgjwu82/h6fkY+1w07z5e/xTsceyqYvm+rQBpA6b/ljnST0OP5YJ4okD3reEowf3TBroqmGSkJPgASSYTrQZClTLJOVgQKaQCAYH+N2qKVapA7G5z73ufHsXiA4r/ns7x7IeN3UfYAHX/4yAM/e/WsAetrDdDZ3IckSH/mzJSy++zIkWeKufyxn9//u4w8/3UVHk59Qe1h04xEIBJNO3/PWHy+4h/89sHXAeUswcRzsbiacjBDXYjitWZwIHccf8+OyOrko5+KpDk8gEJwhPNPdzFpcxLFdLThx8mDJV7k097KpDmtQDgcPpf+dbcumK97F0ZDZ9XOBb+EURXX+ktAS6IaObujIkkxSS6LpGrIk47CMvluoQCA4PcYlUW/evHnY9RUVFePZvUBwXvNg2UMjLpckmF02g7urbmfJPYt6u/HotO3vxOG184lHbmd22YyhmxYIBALBBPKVpX+Du8uLLeqg4XB9evlQ5zPB+ChyzcBhsROIdROMdxOId+OyOilyzRh2O0M3SEQSxMOJMxSpQCCYbErvu4I7vnEz6574M26+8uazNo1yWtZ0ouEo7ZF2OmOdhJNhAKY7C6c4svMTq2xFAjRDQ9O1tLA2khvQMAwz/VMfqlCzQCAYDeNyqC1dOnzdlJHWCwQXMjfNuplnm/6PA4H96WVzPHO5adbN6dfufBcf+7tbM7Z79uHfcfKDDiRZ4tM/WT1gvUAgEEwGR7Yf5/C/dXJTz53s/UgjhxZ9AMBC38KM85Zg4vDYPDiULHoSQWyKFVXXsCsOPDbPkNvomk7noS4wzFo4dpdIAxUIzgeKLjv70+rD/giJgIreDRFvhJgWw67YKMgqGPa8JTh9FFkBXUIzNGQkQEKSpIw6av1JNygAZGX4sQKBYHgmJIl669at3HfffSxfvpz77ruPH/zgBxOxW4HgvEY1VADyHPkA2BU7mqGllw/FtIvM8YZucOLdk5MbpEAgEPTiyncS74kDUNg+A4tkwW01u0iOdN4SnB6GYSBJEvOz5zPPO5853rmouooxWJeaXmRFxmIzn5eqcRVd089UuAKB4AJHlmV0TceiWHGoWeQ58nFbPYA07HlLcPoYhoGMjE22YlPs6UYVwx3vVJdPQDjUBIJxMu42HmvWrKGhoYHy8nLmz5+PYRh89atfZcuWLTz//PMTEaNAcF5ila1U3fhdIokwMS3ONOc0VEMd0aI9Z/lMwv4os5YUMv3i/DMUrUAguNDJnePDW+gme7aXssWX8+BH/wq7Yh/VeUtwekiSxFzPXGa6ZiLLMoqkpEW24bB7bFgcFmxZlhHHCgSCcw/d0GnqOMCRyGFWzF051eGksWRZUKwKsz2zmOeei9OVhSRJozpvCU4PSZKwKTasihWp15020vGWJAlZkcwGBuL3IhCMi3EJao8++ij33XcfW7ZsGbBu8+bNPProozz0kKirIhAMhVW2ku3wkZ16LY18UzprcRGzFhdNbmACgUDQD0mW+Minl9L49DvseqKVgw3HKLv/SoqvmTPVoZ3XSJKEVbFmvB4Jpy9rMkMSCARTSNfRbjb/4Edk7fXw9u0v4fyYi2tmXDvVYQFgtVtw57sIdYYJd0aJueO481yiA/EkI0kSUp+CyqOZJxSrSPMUCCaCcaV8Zmdnc++99w66bu3atcLaKxAIBALBeULzm0doqHoV/+EAWlLHfzhA/SOv0PzmkakOTSAQCC4YPtx5gOyd+dgSdmbvL+aH7zzGgcB+2iJtUx0a8VCC9gOdBFt6UBMqwZYe2g90Eg+JBikCgeD8ZFyC2kjqd05Oznh2LxCcswRbe3jlf95i19a97Nq6l33P7x95o14S2qmLjl/9XcNkhCcQCARjZsdTe8yOwqlnZQYgQeNT79AV65rCyC48dEPURRMILlSqYt9CU1Q0RQUJTkZP8uDLX+azv3tgqkMj1BkmGUviznfh8Nhx57tIxpL0dITQdG2qw7ugEMYWgeDMMC5B7cCBA+NaLxCcjwRbe3jmwd9w1Z+XsuSeRSy5ZxHB1h52bd077HZvHH+df3zz7/nT5z5JKGGOP767dcT3Or5n+DECgUAwEXQfD54S01IY0H60k883rCOpJackrguJrpifYz1HOR46NuptknEVNS6aRggE5wt/fc2X2Fb+Cg33/YK9V+9IL3+wbOrL7CRCSSx2s6KQYRjEwgnChDjSdoSTEXG9OtkYhkFSTxLX4iM2C0pEkultdE0/5wW4hoYGcnJyaG5unrT3aG5upqKigtWrV1NRUcGmTZvYtGkTAIFAIP3vuro6Vq9ejSRJ5OTkUFFRQSAQGLC/1JiSkpL0toJzj3EJavfddx+33347L730EsFgEIBgMMjWrVtZvnw5999//4QEKRCcS+zaupdLb1+QUS9i6b2X8/YTO4fd7j3/Praf3E5Mi/Hg//4N+99tGnKsYRjUffnX/Lzi/3jhn1875ydBgUBw9pM90wv9jOkGBj3Z3UTUCLvahz/HCcZPKBkipsXojnfzYdeHRJKRIcdqSY3Og10EjnYT6YqewSgFAsFkctOsm8m5zItqP/UQY4FvITfNunnqgurF5raixlUSkSTB1hBRf5RgIISUJRFVo2iGcKlNNpquoesacTU2pCuwo9nPE39aS9v+DtSEhpbUMbRz+14iNzeX4uJicnNzJ2X/mzZtYvXq1axevZra2lqqq6tZv34969ato7KyktWrV9PZ2QnAqlWrqK2tpbS0lNzcXKqrq/H5fAP2WVtbS3l5OTt27GD9+vWTErdg8hmXoLZ06VK++tWvsnbtWnJyclAUhZycHNauXcuGDRtYsmTJBIUpEJw7NL12BG+hJ2NZSlw7tqtlyO0uzr0UAEVScH7o5Z3CRoBBa2JIkoQr3wlANBCj60j3hMQuEAgEQ1F2/5UDHGoSEs2L3uOS3EuxyONuHC4YAbtiJ6EmCCXDHAkeoT3aTlyNkdQHugNli5x+2JKMCoeaQHC+kHIeLfAt5AuL/4oFvoUAJPSpr1PmznNhdVgJd4aJBWNEA1Fsuo2sXAdOqwtdF+nqk4mBgYSEamgkdRVVV9ENfUCZgAOvHELXdJpeP5Ke18/1h/OlpaXs2LFjUOFqvNTU1LBx40ZeeOEFysvLM9b5fD6qqqoGdcZVVFTQ3NxMY2PjoPsNBAKsXLlyUmIWnDnGffVbXl7OgQMHaGxsZMeOHRQXF7NixYqJiE0gOOeIhxIkwgm8090D1tlcNjqa/cxaMniHzk3bNgJQcHAGLfOOEOnMZSYL+OzvHuDZu389YPycspmoUZWZiwtF9ySBQDDpFF8zh5UbbqTxqXfwH+5Ctihc+rEF3HvfHeS78qc6vAuCzmgnYTUCGPgc2bSET+CP+XFZnVyUc3HGWEmSsLts6LqBzWnFMIxRdX4TCARnN1bZStWN38UiWZAkiasLr6a2sY6vvfYwm258FFkal19iXNjdNgoW5NHTYaezOYBnug3f3Gw8+a4pjetCIaEl0AwNCQmbYiOpJ9EMDVmScVgcgCmcNb16CIDm1w6z7JNXIMsykizmh8FIpXkO5TJLUVlZSVNTZnbRunXr0ttWV1cP2KampoZ169ZNdMiCM8yEndlKS0tZu3Zthpi2devWidq9QHBOEDzZM+Q6h8dGrCc+5PoHyx7CErdijduIesIZywdj0UcvYtEfXczBt47y1Of/j2cefE502xMIBJNK8TVzWPXvd/LAU/fzmdr7ue4vlwsx7QwyyzObbLsX3TBwWpzohoHL6qTINWPQ8Z7pbhweO7FgnM6DXXQd7Rbd9gSC8wCrbEWSJN755fv89AtbifyXwf7OD3n1+CtTHRp2t438eblcdPN85i6bRXaBR4hpZwirbEWRFID0MZclGatsTY/pPNhFqMMsFxBqDxM4GsQwDDRVR02o6NqZdxE2NjZSVlZGTk4OjY2N1NXVUVdXR0VFBRUVFYBZI62mpiadetm3JllzczMrV64kJyeHhoaGYfdZWVlJZWXlqGOrqqoCYM2aNcOOG2r9unXrqKmpGXRdU1PTqNxpdXV1lJWVIUkSK1euTH/2VN24kpIS6urqAFOkS/1UVFRkuOMaGhrS+2lubk4fy6HiE4yOMTnUDh06RG5uLl6vF4AXX3xxyLGBQICNGzdyzz33jC9CgeA8IhEe+kbmplk389qRt/nDxa8OWD4YzW8eof6RV5i1pIjln1rM0cYW6h95hZUbbqT4mjkTGbZAIBBkYHWI9M6pwGPz4I85UHWVUCKEqqvYFQcem2fQ8fFQgmBrDzanFbsni2QkSbC1B2+hRzibBYLzgBPvtGL127EChSdm4Y/6pzqkNMLxdOZRZAXVUDEw0HWNwPEgoSORjJIMh/9wDEmWMHQDSZbY/b/7mLNsJsgS6AaGAQUlOeTOzTljcafSNSVJorq6mqqqKnw+H6tWrSInJwefz8d9992XTrdcvXo1lZWVaddXcXEx9fX1GS7skfa5fPlyVq1aNWJsDQ0N+Hy+EYWvVOpnfyoqKtICV183WkNDA6tXrx7N4WHVqlWUl5eTk5NDZWVlOpby8vK0C664uJiamhqamprScTQ3N1NSUkJTUxPFxcWUl5fzwgsvkJOTQ11dHevXr6eyspIdO3YM8+6CkRjTFXFpaSklJSVs27YNMH+5gUBgyD+w7m5R10lwYWF3DX2DEusZ3hVweNcxwnOCLPAt5La5t/PayTcBs1aGVbIOGL+z9l1mLSniY9+8FUmSuOLjl/LcN19kZ+27QlATCARTQne8G7fNnX5CLphYDMPAwGCudy4+u4+TkZOEkqEh0zkjXVFsTivZM8wHofiy6D4RJNIVFYKaQHAecHH5Ag69fQxrscyXr/4Kly68eOSNphhVV0XNzUkiVQfNpthQJIV3n9rG4beODz1eN2h67TBNrx3OWD7v6lnc/vDNkxnqoKSEq77awrJly2hoaMgQq5YvX87TTz896Paj3ee2bdtGJag1NzdTXFw8ps/Rl9LSUkpLS6murs4Q1Orr6wcV4IbC5/NRXl6ebmSQIi8vLyO+lEMPTKHR5/PR2NiYHtP/GI0lBsHgjMl/W1tbm2EJXLZsGbqu4/f7B/1Zu3bthAcsEJzN2N12YHAnWiKcwDaM4BY4GOQfV/8D/3zTvyJLMoF4AICdJ4coZHmsm9mlRUiSRKg9zEv/9gbTL80ncEwI2QKB4MwR9kf4xX/+mo3/tYlP/+ZPeb/zvakO6bxFkiTmeuYy1zuXhBZHliQcsn3IznlaUsPqNB/IJKNJetpCWJ1WtKTotCcQnA/MKZvBff/zcW5ecz3v/mQ/j636GbVf+tVZVwIkGVc5dvg4B9sPcjh4eECRfMHEIEkSNtmGXbGjSArXf3E5866dNaZ9FF87h+s/95FJinBkli9fnvHa5/OxbNmyAcvGu8/RUlxcjN8/PudnKvUylX4ZCATIy8s7rf301WIaGxspLS1Nv163bl3abRYIBNLvN1j8fbcTjI8xCWorVqxg6dKl6deDFdfrSyrnWSC4ULC7bdhcNmJD1KiZtXjwhgT7nt9PsLWHHT95h7d/vJPgr2IUvme6zF547FV2PvPugG18s7I52tjC4e3HePqLz7L/5YO899sD+GZlT9wHEggEgmGIBmI8VfF/nGzowvmaD1R448TrUx3WeU1/J5ohGekHMP1RrArJSJKwP0LgeJBYME6oLYRiFQ5CgeB8QFZk/EcC1D/yCv7DAbSkjv+w+fpsEdXi4QSBo92Ee6L0+EMY6ESSkakO67yl7xxhdVm58cGruPaLy1CsQzcekGQJxSpzw+c/wg1f+Ag259nlYJ7KLpjl5eUEAoGMmm1Dkapj1p+UMy2lnQzVjCBV36zvT98aaKtWrcLn86Xfp6GhYUDX0VS9tVSduNzc3EFjGmq5YOyMq0Lk/PnzByx74YUX2LVrF0CG+CYQXCiUXD+HYGtmc4LU66E6fF52+0Ju/MLVXP3pUq7+dCn3fvHjWC43bdt/VVnB0nsvH7DN0tWXc2xXC7v/dx+pTteRrihX/NHZb/cXCATnB1k+B7NKzYL41oSd4vBF5GcVTHFUFwa5jjysspX8rALyHIM/6XbmZJGIJIl2RYn1xAm1hwl3RnHmZJ3haAUCwWSx46k9IAG914IYgATbf757CqM6hS3LimyRcVqy0OI6dskuOg6fASRJwipbkWWZy8oXsurf7kS2DH7rL1tk7v7uHZRcPxcMCfHrOUVKmNqyZcuw45qbm4dd37c5QWdn56Ai4Y4dO8zSDn1++jvJ1qxZM6SpadOmTVRWVlJbW0t1dbVwoZ0hxiWoPfzwwwOWFRcXk52dzTPPPCO6fAouSJbcs4iDb2Q+Fdz3/H5u+MJV6dfxUIJf/V0D7U2dg+5DkiRuK7wDYMhi08XXzGHlhhtJRlUMXcfhtXPLl6/loltLJuiTjI7WcAv/ves/eWZ/Hc/sr+O3h34zpu1/e+g3PL73h5MUnUAgmGzK7ruCK+++lI/95838858/yh8vFM2IzgSKrDDHY9ZSG+rm1O624S30oNgs6JqBI9tB0eXTRf00geA8ovt48JSYlsKAruNnRwkQSZZw5TrJK8xl0WWXMjt7Di6ra6rDuiCQJRmbbEORFSRFRksMUR4goZnziCRhzbJgsYkadymKi4uprq6msrJyWJdaXV3dsDXZUpl7q1evHpCCOhYqKirSHU/7v1+qUUPfmmqpdM9AIJBRX00wcYzr25IqfNiXlGtt/vz5PProo+PZvUBwTuIt9FC+/gbeeqKRaQvyCJ4M4fDYuez2hekx8VCc9gN+4oM0Kgi29rDv+f1pUa6+6hVmLinK2D5F8TVzprQBQWu4ha+8/CU2r/whbpsbgMf3/pBn9tdx78KhJ5XWcAvP7Dftyq8df5Xb591xRuIVCAQTT35xLvnFInVgKhiNy8PutmF328iZLcoBCATnI9kzvfgPBTKWGRiEvUHCyfBZIV45vPapDuGCJTVPHHzjCJIEhkFGl09DN5AkONp4giX3LJriaAcyWLrlcCmYo0nNHG0KZ4p169YRCARYsWIFmzdvHuD8qqmpGZB62Z/S0lKKi4tpaGigtrZ21O891H5qa2sHTRvt+7lSDRUCgQB+v1+keU4SYxLUuru7OXjwYPr1wYMH2b1796DC2vbt26mvr+ehhx4af5QCwTlGQUkeBSVDF5v0Fnp44GdrhlyXSv3sS1JLcjJyklmesRUXnUye2V/H7fPuSItpAKsWruFPnrtvWEGt0FXEF5f8PwAOdO2f9DgFAoHgQkDTNQwM0UFPILiAKLv/SuofeSVjmYTE4vsuw2lxTlFUgrONA68dSpeIKVo0jasfKOXNHzXS8s5JDAOaXjs8ZYJaY2Mj1dXVBAIBNm7ciN/vZ82aNWzcuJHt27fT3NxMZWUlVVVVbNq0ierqapqbm6moqEg7x/puD6ecZUPtc/v27cDINeFTrF+/nlWrVqW7Yvp8vnRjgXXr1o2qzltVVRXbtm07jSOUSSoNtT/19fXpY1NaWkpubi61tbXp8aWlpaxevRqAtWvXUl5eLrp8TgCSMZgaNgTd3d1s37493e1TkqRBxTQwi+rV1NSc1XXUgsEg2dnZdHd34/V6pzocgWBIdrY1UrPn+4STYZwWF59b/HmWTBv6u7X/1UO8/UQjsUCM7Jleyu6/clKcbJ/89Ro+vegB7pj30YzlH//Fnfzjtd8eNsYUX3npr1k8bQl/segvJzw+wYXJRJ3bxRxxehi6QaCjG6vPkiG2CyYPwzDoTnTjj3YiSTIGBoXOQpzWoW+mo4EY3a1BDA1sbivuPJdIBRVcEEzkuf1smSea3zzC9id303W0m6wcB9esXcbC6+ZNWTwjYegGqqZisVhEPbVeYrEYBw8eZP78+Tgcjgndd097mJ999n+RZImlf7KIyz6+EIfVgYzM7v/dxx9+ugtDN/jUY3+Mu8B0NOqajpbUMHQDWZGRLTKyMq5qUQLBlDPa79lYzu1jeoSZnZ3NihUrWLFiBatWraKuro7vf//7Y9mFQCAYI4Zh8NT7P+d46DgAgXiAJ/Y+jtvmxmvLZppzWsb4bT/fTeNT76Rfp7o9rdxw44SKaqFEiHAyTKGzcMA6l9VFU3fTqAQ1gUBwfmAYBo0v7ubtp3YSi8coWO/izy7/86kO64JAR6c90kZSS9IZ8xPXYiiSQqE0HVlWsMrWjPGB40GO72lFTaj4ZnqJ9cSIdsUoWJAnRDWB4BwkVQJES2pndRdfQzfwd3bR1taOZlcpnjt/WOFfMDFIEswqLeLK1ZfgW+Alrsex6BasipUr77mUosuns+Pp3uYWmGJaIpJEV/V0Wqiu6VjsFiGqCQT9OO2cgPLy8iGfKLz44ovk5uayZMmS0929QCDoRZIk3vPvy1jW1H2AB1/+MgDP3v3rjHUfNDRl7qC321PjU+9MqKB2MtI65DqP1UNPIjhh7yUQCM5+JEnicP0JlDYrLqzse+U9jEWGcB+cARRJIZQMcSJ0AkVWyHXk0BI+gT/mx2V1clFOZvfn4Mke1FiSLF8WikXG6csi1BEm1BkWgppAcA4zlJhmGGfHudgwDML+CNFEFJLQEw0KQe0M4M53seJr15HQEqi6ilW2kNSTaIaGLMlMvzifj/3drenxmqqja3paPJMVGV3T0VVdCGoCQT/G9Y1YsWLFoMvLysrIzs4WTQkEggniwbLBaxEOtjwaiA4caEDgDHd7CifDZ/T9BALB1POR+01XanBaF+5cJ19oqGBX284pjurCYIFvIXO9c/HasnFbPeiGgcvqpMg1Y8BYxaJgsVuwOa3Y3WaxcIvdQiKUPNNhCwSCSeR4y3H+5/f/Q/We7011KIApzOTk+5CQkB0Sx0MniCQjUx3WBYFVtmJVrFhkC5Ze17IsyQMczGA6CRVLr5hmMbt/SrKEro+6UpRAcMEw7qq1hw4dorGxMd2SNUUgEJiQonsCgQBumnUzzzb9HwcCpwr4L/AtZFHe5ZwInWCG+9QNk29W9oBuT0jm8olkuK5RPcmeCX0vgUBwbjBzcSFL/vYSshe4qN7zfY53Hx82RV0wcXhsHjw2L/5YF6FECFVXsSsObLIVzdBQpFPOFUe2HVvAhjM3K71MjasZrwUCwbmLrum885v3ee2Jt/FP62HHytf4eMknmOGeOdWhYfNaKV4wl4DRzeHgYdqj7UOmpwsmDkVWkA1TJNN7G9hIkoQiK2i6hiKfmiNkWULTQLbKaWejoRsoVuFOEwj6My5BbefOnZSVlaW7WqRasfr9fkpKSsbVElYgEJxCNVTAFNFum3s7vzv8PIZhULVtI8d6jvLl0ge5esY1QJ9uTxLpdE8MWLpqESf2nmTGoukTEpPb6gEGd6KdLW3aBQLBmUWSJP6p5e+h5dSy4VLUBROHYRgYGMz1zsVn9xGIB0jqSY6HjmORLRS6irApZjqnO89FtCtGuDOCxW5BjatYHVac2VmoCQ2L7eytwSQQCEZGjWu8s/V9LAkr047NZObReRzqPnRWCGqHQ4cIJyPEtRhOa9aw6emCiSPVSNCm2FAkBc3QALM7dCr90ypbkSQJ2WKmeBqaDr011CS5d7luIEmcFSnEAsHZwLgEtZqaGpqampg/fz47d5opHamungcPHiQQCIw7QIFAYNq0q278LhbJ7IZ0+7w7eOqDn/Pz958E4LF3N1M6vQybYqP4mjms3HAjjU+9Q+B4N76Z2ZTedzlHtp/ggxebuH7dchZ9bPwXLG6bG5fVNaQbbUmBaEggEFyIPFj2EP+y41EKD81i4a7LcQW9hL1BLrm3ZKpDO6+RJIm5nrnpm5xsezbHe44R1aIk9AT+mJ9Cl9lExu62UbAgj1BnmEQoiTM3C2dOFrFAjEhXFG+hG2uWcIoIBOcqNqeVaz9TRv2mV1HKdL55/zcoKDg7HMJFrhm0hE8QVaPYEja6W0PIuhVHnou4NSHqOE4SkiRhk23pOcKCBcMwSGgJAHRDRzM0LJLZeMBit6CrOnqvM022mG41NakhSWa9PiGqCQTjFNRKS0uZP38+AMXFxWzYsIHvfc/M0Z8/fz4vvvji+CMUCAQAGTZ4SZL4RMndHO05wlstb1K5/OG08wBOdXtK0fzGET54wWxW8OYPdzCnbCae6e5xx3T9zBtoDbdkLEu9Fh0+BYILk5tm3cyLv/09M19aYKaUIOHp8nHiMT/N+UcmtDmKIJP+NzfTnNNpCbdgoFPgLMhYZ3fbMm5cQ+1h1ITpWOhpD5MzO1vcLAkE5zDzr53Dqn+7k7z5OVMdSgYemwd/zEE8lCByLEa4O4rVaUczdNpDnaLb8CTS/5wuSRIW2YKqq8i9/04hK/KABgRqXAUDDAN0VT+rO8oKBGeKcSVC9/1SZmdns23bNg4fPpxe1tjYOJ7dCwSCYXBanXx1WSX/evO/szBn4bBj518zmyWrFoEEt3zlugkR0wDuXbiKN46/nrHst4d+wxeW/FX6dSgR4huvf42mwIFB9xFOhkUDA4HgPEI1VPLempEW0wDz/73dhgVnDqtiZZZnFjNcMzNqqA2GMzcLm9OKrMh4Cz1CTBMIznEkSTrrxDQ4lZ6en5yGLZCFgkJ3pBt7to1kLEmoU1wTnkkUWeltVjCyK9l0pYEkg2wR9dQEAhinQ80wDDZs2MALL7zAtm3b2LBhA+Xl5dTU1NDV1SWaEggEk4wkScz1zhuwfPOeavKz8rl7wT1IkoQkSVz1Z0tZcMM88uZN3MVVoauIyo88zON7f8hC30W0Rlrx2LzcMe+j6TGhZA8HAvvpSZxKDQ0lQtTt30I4GaY10krP8VfT+7t34aoJi08gEJx5rLIVV7cbDT1zxRR0GxaYXdz6uwx0Q6c13ILPnoPT6jTHKTLeIg9aUhc11ASC85REIsGezt0sK1o+ZTGk0tNbaSfhSiCZpie6urvw2rNFt+EpQJYGimO6oaPqKhbZkl4vyRJK7/wgHroIBCbjEtTWrl3L5s2bKSkx66KsWrWK5uZmVqxYgSRJ1NfXT0iQAoFg9Lx09EV+2fwsAM3dzfzNsq+m1/UX05rfPMLbTzQS7oyQM9vH0tWXjzkdq8S3gBLfgiHXF7qK+PmdWzKWuW1u/mLRXwLwxSX/b0zvJxAIzn6yZ3rPSLdhwenREe0gokaIqBEKsqaRbTd/L5IkDRDT4qEEXce6UawyFpsFZ06WSMcSCM5BXn7tFXb/4D32X7wX62dsLC5YPGWxSJKEzW3F4/QQD8fx5njJy80j2hkT3YbPAgzDIKknzf9rCayK7ZSoNoiQpms6WlIDSUKSBk8XFQjOV8b9l7527VrWrl2bfr1+/Xp0XUfTNG699dbx7l4gEIyR9khb+t9X9l4s7WrbyRcaPseutp3pdc1vHqH+kVcItoaw2MwCpPWPvELzm0fOeMwCgeD8ouz+KzMX9HYbLrvviimJR3AKwzBQdbNztISMw+IAIJKM0NzdTCQZSY+NhxK0fdhBtCtq3ixhEGztIR5KTEXoAoHgNAkc6+aD7x7B4XdxyY4l/ODNx/h8fUXGdeGZxp3nwulxMt07nWxbNtHOGFaHFXee6BI/1RgYp15IUrp8g6ZrxNU4mq6lV+uaTjKaRFdPudK1pI6u9XOpCwTnKeNyqE0WDQ0N6fprTU1NlJSUsH79+owxjY2NNDQ0ANDZ2UleXt6AMQLBhciai+9nfnYxe9r3sLhgCQcC+3li7+McCx3lib2P47a58dqyadzyDlaHhWRMJR5KsODGedicVnbWviuKhgsEgnExWLfhsvuvYL44t0w5kiQxwz2Drpgfi2xFlmTiaoz2aDvHeo6hSAqF0nRkWaGnPYyW1HAXmDe47mluIp0RIl1R4VITCM4hfLOyWXDzPA68fAg1J4EUkzmuH8u4LpzmPLNdQAfrNuzOc4lzy1mALMnYZBuqrqLICgaG+TDGUEnoCZBAMnpFtqQprskWBQmzzlpKUBMuNcGFwKQKap///OfTXT9HS2NjI4FAIEMcKykpoampierqagCam5tpaGjIGNPY2Mjq1aupra2dmOAFgnOY5YUfYXnhR/j4L+7MWN7UfYAHX/4yAJ84/qdcefelHNl2HEmWWPSxizB0nW1P7p6CiAVjoTXcwjP76yh0FQHgsroy6taNZ7tdbTvZ1b7TrG8XbuW6mdePat8CQX/6dxsGiPXE2fbkbj7yZ0uwu8RN01SS48gF4MOuDwgnI8S1GFkWBx92vU97pB2v3UOuUUDOXB9qVMXusWG1W7A6rUT80SmOXiAQjJVrH1jG/8ZrOXJRE8imA6nvdeGzd//6jMfUv9swmJ0k46EErjznGY9HcApJkrAqZqOCmBpDN3R0Q0dCIq7GUWUVRVJQJAuKVcEwzEYFkiQhyRK6ZozwDgLB+cGYBLVHH3101GM7OzvZsmXLmAW16upqGhoaWLXqVGHyVKODlKBWVVVFZWVlxnalpaUEAoExvZdAcL7zYNlD/MuOgd/bK/MX45uVTdsHndy96Q5iwTiSLHG0sSVd4+jI9uPkzvWlnQmCs4PWcAtfeflLbF75Q9w2s1vr43t/yDP764Zt6DCa7Xa17aSpuyld3y6UCPHgy1+iKXBA1LoTjJvdu9/h9Ud3oAQtxHvirHjoelHU+CygyDWDlvAJomoUGYOeRAiLZGGGewaKVQHdIGdOdvp3lYwkzeVALBjH7rYhyeL3KBCc7WT5HKz61N2DXhc+WPbQFEQ0EL/fT8vxVnIdecgWmaxsx1SHdN6ixlUs9tFJAVbZSlJPohs6BgaaoSEZYFNsSDqAhMUqp+cCQzeQpN6OrrpxRpxqDQ0NrF69mh07dlBcXDwp79Hc3ExVVRV+v5/c3Nx0Hfn169cTCASoqalh/fr11NXV8fTTT1NXV4fP52PNmjVUVVXh8/ky9rd69Wrq6uooLi6moqJCZNtBWs/pf6z609zcTGVlJQ0NDTz88MNTeuzGJKh95zvfITc3N+MDNjY2UlxcPOBDNzc3p//IxsLKlStHHOP3+6mqqkoLbH2XCwSCU9w062aebfo/DgT2ZyxfMm0pS1dfTv0jr/Dbb7/M7NIijja2cGxXC7dtuJFwZ4SGR1/DMAyWf2oxV3780hHfaywTs+D0eWZ/HbfPuyMtigGsWriGP3nuvmEFtdFs99tDv2HDR/42vd5tc3PPRav4n13/xb0LV6WdbQLBWGkKNPGdPd/ihtgdKMCx3S2EOiJ4hGA/5XhsHvwxB6qu0hXvRNU1LLKNXEcuUo5MsLWHnpMhrE4ryUiSRCSJt9BDPJygpy1E2C/jmebC5hSOQ4HgbGew68LZnjncNOvmqQuql0A8wMnoSeJqgp54D44ehxDUJokTe0/y3Ddf5M5vrqBo0cipvoqsoBpqr0tNw8BAQsYiWTAUAy2po6k6kixh6KaIplhldFVH1wx0zUCxyJP68CU3N5fi4mJyc3MnZf+bNm3i6aefpqqqivLy8vTyQCBAZWUljY2NlJaWAmajxlWrVlFWVkYgEBigWaSora1l5cqV1NbWjiggXShs376d3Nzc9LEciuLiYmpra09Lb5poxiQXl5eXc+DAAbZv38727dt5+OGHaWpqyliW+qmtreWRRx4Zc0CrVq0a8Ee3ZcsWqqqq0q8rKiqoqalh9erVaRVz06ZNVFRUDLvveDxOMBjM+BEIzmdUwyw8vcC3kIorPk9+VgF2xc4d8z+arnGUSsGK9sS4bcONzL9mDtt/vodkNIkaU+k60j3i+5zYe5LH/7SWlr1tI44VjI/Xjr86QNhKiWTDFRcezXZvnHidx/f+MGPMQt9Cc0z7rnHFfS4g5ojJozi7mDlzZrPnurcJFvm58TvLhZh2lmAYBgYGc71zuWbGdVyUcxFOaxZ2xY7dbcNb6EHXDHo6Q+iagbfQg91tI9xpNi9I3TAJBBcC5/o80fe68M9nPUDZK9fTecJPc3fTFEcGHqsHm8uGzWtBchlkF3mmOqTzlm0/2YWW0PjDT3eNarxhmOd4u8WOy+rGrtjTXT9lRUaxyoCBrprNaxSrmfqZmhsMffLniNLSUnbs2DEpwlRNTQ0bN27khRdeyBDTwHRSVVVV0dzcPGC7iooKmpub07Xh+xMIBFi5cqUQ0/ow1vJdZ8OxG5Og1lfUAujq6mL+/PmDjl2xYgUHDx487cDq6uqorKxMq7Z9bXzl5eVUVVVRV1dHTk4Oq1evpry8nHXr1g27z40bN5KdnZ3+mT179mnHJxCcC1hlK1U3fpd/vulfubPkj/jBbT/iyY89hdtqCinF18zh3n/5GDOqsvnVbU+zv+g9NEPj6r8o5bKPXkSWz8FVn146YL9xLZ7xeqwTs+D0CCVChJNhCp2FA9a5rC6ahrggHu1218647oJ2oYk5YvKQJIlPL/oL/uiuj3HTN67i0Q+rprS7nOAUkiQx1zOXud655GblsmTaEkqnlaVTPO1uGzmzszEKVUK+bjS7eUPuLfRgzbJic1pxeOxT+REEgjPGuT5PpK4LH8rfQGdVlMKm2Sx6s4x/3f4vaIY28g4mEUVWyM/KZ97secycNZNDoUMZXYcFE8OJd1ppfa8dgNZ9bZx4p3XEbSRJwibbsCt2rIoVl9VFliUrPU+YopqCbtExFB1kkGSpV1gD2SKds6UBmpubqaioGDRlsy/9y1EBaW1iKIdaTU3NiPrFhURDQwM1NTVTHcaYGVN+Vn/xbKSaZeOxXKaskilhrba2NiMfetWqVWzbto3m5mbq6uoA2Lx587B/6A8//DAPPvhg+nUwGDznJkKBYKxYZWv635IkYVMy03KiapSfv/8kgXiA/9r1HxT7ilngW8gNn/vIoIXD33jzLX667Sd84ZOf4/KCKwadmGdcMVC4EYyfk5GhL3o8Vg89icGflI92u77pnin296aFLClYMoZIz03EHDG5THNOx2Fx8N87/yvddTiyJ8Hh/2sh3BIhe6aXsvuvFF2Gp4D+tez6v46pMULJEGCeT+Z652GxKfhmetE1fcD+YsE4FrsiygAIzjvOh3nCKlvJL87D6XES6Yzi68zj9tm3oEjKVIeG0+pE1zVaIyc51nMMPWLgirjQYjpZbtEFdCLY9uTudGqmJEts+9kePrFx5Ov2keYJrTcVVDMMDF3FptiQlaHTPDVVR1YkJEmisbGRtWvX0tzczAsvvJB2e9XX1wOnaqw3NzcTCATYtm1bxn1/SvRKZcmVl5cPuc9t27YBA41CQ5Eat2bNmmHHrVmzho0bNw5Yvm7duoxa8H1pamoalcNqvMcnRUNDA42Njfh8Pnbs2EFFRUVGamVfMav/+ok6nnV1dekadH6/nx07drB69WrglDtt48aNad2n735TNexKSkrw+XyTlt47VsZ1pXPgwAF6enrweAa35P7hD3/gnnvuGc9bpIWzsrIyDh48iM/no7GxkY0bN6YP+qZNm9K5y01NQ1uW7XY7drt4iioQ9CWqRrk45xLebn2La2dcx4LeFD9ggJimJTV2VL/L4q5r+N2hV1n4rYtPe2IWTDzhZHjCt9v6YR2fXvTABeFcE3PE5PLZ3z2Q8Tq8O8Y7L+3vrcUi4T8coP6RV1i54UYhqp2F2GQ7CT1uFguXTiU49C82rSY0etpN8c3pyxKd+gTnFefLPGFzWrl+7XI+3NbE4j+5jML86VMdEgAHu5vTXYdtSRsHPzyEFFewaVZm+mYS7YpRsCBPiGqnSd+H4GCmYk7Uw3Cp9z8DA4t8SmIYrPmQrunoqo6hmZ1BU+makiRRXV2ddoOtWrWKnJwcfD4f9913XzrdcvXq1VRWVqZFquLiYurr6zPea6R9Ll++PKMJ4lA0NDTg8/lGFL5SqZ/9SZWq6u9GSzVRGA3jPT5gClkbN25kx44d6WU5OTnpJg41NTU0NTWlP0OqHn5TUxPFxcUTcjybm5upr6/PiGvTpk2AmYGYakT58MMPD6ih1tjYmG460VdIbWxs5L777hvVcZwsxtVyY/369SxdupQf/OAH7Nq1i0OHDrFr1y4ee+wxFi5cyP333z8hQa5cuTLdOQNg7dq1Gfm169evp6mpCb/ff07aBAWCqSTXkcvXrv4Gj9ywiU8vemDA+meb/o8ToeMA/P63r2HrMgvEJoIJ/rBjG63vtadrI/SdmAUTj8s6dM2pnmTPhG/3yB++w+JpS4ZtdiAQjJb+XeQW7rw8LaYBYAASND71zpkPTjAsDouD2Z7ZFDoL8dq8Get0Q8cf86fTxaKBqPm77FMyJx5K0HW0m45mP11Hu4mHEmcweoFAMBjzr5nD7X99C5H9cWq/9CseW/Uzar/0K5rfPDJlMRW5ZuCyOtENAyVoIRqMoakqPnsOhgHxUJxQ5+k9PBSccqf1JfUwfLwosoJNsWGVrRkPXcCcJzRdTddi01XT2Wz0mSd0Tcfn8+HxeHE73Wn387Jly2hoaMgQWJYvX8727dsHxDCY6JUSw/quW7ZsWdpZNRLNzc3jckKVlpZSWlo6wKFWX18/oB7bSAz1WUZzfNauXcvDDz+csSzVfTRFQ0ND+t+pppP967+N53g2NjYOqDU3GlETTomEfd+3uLh40jq6joVxCWrFxcVs2bKFjRs3UlpaSklJCaWlpVRVVfH973+fJUuWjHmfOTk5aaUyReqPuKmpacg/6uLiYh5++OEM1VUgEIyey/IWUdTPhbS/60Mee6eGL77weX723k/5nfdX7LzldSKuMHuv286uH7036L5e/f4fzkTIFxxuq+kGHsxRFk6GhxTOTme73x76DR6bhy8u+X/jCVkgSHPTrJszHLCuoPeUmJbCgMDxkRuhCM48kiThtnkGuA264934Y50cCR4mlAjhLnDhzM1CsSk4c7KIhxIEW3uQFQlnbhayIhFs7RGimkBwFtD85hHqH3kF/+EAWlJPO4Xfe/XDKYnHY/NgV8yuw+3+DoJyN0k9iYEprljsFhKh5JTEdq6Tcqf1bxAwkQ/DJUlCkQemDqu6SlJXSegJdENHsSnIitT7I6NrOlrS/B1/ZPlyQEJL6mmRbdmyZRn7G2sh+uXLl5/29sXFxfj9/jG9X38qKipobGxMi1OBQIC8vLzT2tdgn2Wk49PY2EggEBjg+iorK0sLb+vWrUvrKIFAIB3rYJ/9dI9neXk527dvp6SkhMrKShoaGiguLh5RWGxubqa5uXnA5xzLe08m4xLUwFRdDxw4QFNTE7/73e9oampi//79rFixYsz7StVk6680ppTMsrIyiouLB+2iAeYBLSsrG/P7CgSCwXlmv1mfUDM0wskw+7r2cWLeEV5e9Ut03cDmH7ydeeBYULjUJgG3zY3L6hrSVbakYGADidPZ7vXjrxFOhjPEtFAidJpRCwQmfbvL3XfRJwlnBzHo1/lLAt+s7CmITnA6GIZBIB4AzHnCqliRJAnJA13eTqJalEhXFJvTSvYML1pSxzDM8gGhTlFsXCCYanY8tQckTjlKDTAwePHHr6XdRGeSvl2H502fR46ciyXXQlJJ4s53oms6Nrd15B0JBjCYOy3FRLnUBkM3dHSjt9am0ZsaKkmgQFJKoukauqanYzv1d2ekXWpTKZqUl5cTCARGrB0PpOu696d/c4KhmhGUlZkNgfr+DNUhtC8jHZ+UaNbQ0EBdXV36Jzc3N8OhVldXR1lZWbrBwuk684b6HD6fj4MHD1JeXk5dXR0rV66kpKRkxGOb0n7OBvFsMCasWuz8+fMHNC3YunXrmGqo+Xw+1q1bN0A9ra6uprS0NP2Ht2rVKjZt2pTR+TMQCFBfXz/mVqsCgWBovlT6FeZ65/F2y1vs69ibro1gyAaaRcWQdJLWJLbEwHoifWuppSbHweooCMbG9TNvoDXckrEs9XrJtMEFtbFs1xQ4QE+yJyPNM5QIsbt9F9fNvH7c8QsuXFLd5SySBUmS+O3qBg7XtJ66mev9f9l9V0xxpILRIkkSszyz8Ef9SJKELMnE1Rjt0XaO95YKcESceArMztKJcAJd1VETGop1ajsKCgQC6D4eZOBzDQmpQ+Htlre4esY1ZzSeVNdhSZKIz0lghCAZTZLryyXaHcPqsOLOG7qMhWBw+tdO689E1lLrjyzJ2BQbqq6iSIp5H2EYqIZKUjfdhrIuo1iU3ljMeDAyU0KnisrKSmpqatiyZcuwHTmHMvyk6NucoLOzc1BxaLIy7VJmpfLy8iFTJDdt2kR1dTX19fXjTqMc6nM0NjZmpL+mmkls3Lhx0PpzKWdaKp7RiJpTwZgcaocOHSIYPNVF7sUXXxzyZ+vWrYN2uhiJqqoqGhoaqKiooLKyktWrV6eL4PUd4/P50mNSf+hCTBMIJpYsSxafvORPeGDRX9IUbMpwk0SyQwRzA3Tn+4k6T6USGpJO0pLMsI8f29XCz9b9glf+5y3a93ee8c9xPnHvwlW8cfz1jGW/PfQbvrDkr9KvQ4kQ33j9azQFDoxpu9ZwC7UfbsFj9fD68dfSP0/s+xGFLtFoQjB+rLI1LazfcWc5KytvIG9uDopVJm9uDrdtuJH518zh0NtH2f7z3VPikBCMDatsZbprOtOc0zjY3cw+/3u0hE/gtGbxvv89dvi3s/voLjRVQ5IkYmqMEz0nUJXMtK1wZ4Suo92EOyNo6sDOoQKBYOLJnullYOa9gS1P4aqiq6ckptQcYXfbWHBZMbPnzcJqt+It8mQ0JAh3RoiHRer4aBjOnZZiMl1qKVFNkRUSWoKYGiOpJZEliagaJaJGiCaiZhwSGIZOQk8OcLGrCQ01oQ3aXXqyKC4uprq6msrKymEFnbq6umHrgVVUVABmLbD+KZOTTXl5+aD10OBU3bRUE4O+Yloq3TMQCGTUVztdmpubB9Rpq66uHtaFl5ubm66VNljdvLOBMTnUUnXSUkXnVq1aRSAQGNJ+1919enVQhlN/xzJGIBBMDD9//8m0O60vB67cS9lLN5C0Jog6I8QdUXz+PPYv2cXMQ/N5se41/vSKVRzb2UKoLcx7zx9g5pVFFCw8VTfAMAwM3RjQKU4wOIWuIio/8jCP7/0hC30X0RppxWPzcse8j6bHhJI9HAjspyfRM6btvvLylwgnw7xxIlN4A0QtNcGkUHztXIqvnZux7NDOo9RvehVd1YmHElz7mWUj3ggIzg6KXDNoCZ8gqkaxSBbiWhxPjgd7OIuu1m4sLpnucDedmp9ch4+4GkOWFayylUQkiRpXUeMqjuzMcgKGYQiHs0AwCZTdfyX1j7ySkfYpIbHiMzciSRK72nZSs6eadVdWDOuCnyzsbtuAjp6GYRANxIh0RUEC73Q3dve533l1shjJnZZiMl1qfbHKVpJ60kwD7f2bkxUJ2ei9D5BARet1tMl0Bbro6upK3y8ABLoGpmCmHsCNxsU02hTOFOvWrSMQCLBixQo2b948IJuupqZmxDpgpaWlFBcX09DQMKEmoME+y2DLNm/eTGVlZYbo19zcnDGu779TzrBAIIDf7x82/XMsx7OqqirjWDU3N7Ny5cr069LSUrZv305paSmNjY3peKurq6moqBjQKbW5uZnOzqk1a4xJUKutrc04mMuWLeN3v/vdkOM/97nPnX5kAoHgrOCd9j3s8+8bdF3rnOM0L3qP3NZpuLuziWdF2X7rq5yce4yDV3wAwK3vXE8ypiJbZAzdYObizEm682AXv/pGAzMXF3LpbQuZtaRosLcS9KHEt4AS34Ih1xe6ivj5nVvGvN1g25wJdrXt5Efv/oAHLv/MlFywC84OVF3lsXdq6HotRK46A4BYT3yKoxKMBY/Ngz9mFhSPJCNYZCtZzizmT5vH0ZYjhFsjqHKSnDkejiQP0dneQY7Dx0LfRel9WOwWFEvmA5ZYd5xIIIrNZcPpc6BYBxa9FpzfRJIROqId5Gfl47Q6pzqc84bia+awcsONND71DoHj3WTP8LLsk1fiWuzgQGA/T+x9nGOhozyx93HcNjdeWzbTnNOmLF5N12gJt6B1GzhxggG6LpzMw5Fyp/VvRjAYKZdaqmTLZKDICqqhmqmfvct279nDT37wEwKBAI9UPcLJ9pPcs+Ye/v4fHmXH9h00NzdTWVnJd771Hb776KM89sPN6XTBlHPse//zfQKBABu/sxHDMCgpKaG6utpctnEjfr+fNWvWsHHjRrZv3552O/XvvjkU69evZ9WqVenURJ/Pl24ssG7dulHV96qqqhp1d9G+NDY2DvtZUsenqqoqnbrZ9/gUFxezatUqiouLqaioSNecz83NTQtW9fX16e1KS0vJzc2ltrY2XU+tvLycioqKcR/P1atXs2nTpvTxCgQCGWW8UsJfIBDIcMuVl5en41m+fDl+vz/tXKupqaGrq4v//N5/YpEsgzbGmEwkYxz5FAcPHhxQN60vO3fuZOnSs/fmKBgMkp2dTXd3N16vd+QNBIILkA2vfJX3u95PFxR1BTzM23cRxxYexMCg7KXr6Zrewa6b3hywrZK08Mndn2XNt+8iGVN5a8db/Dz604wnnbue2cvbP94JwHXrlnP5nRdn7ENNaFhs4ubpfKQt0kYw0c33d/8PH3Z9yEU5F/G5xV8Y9wX7RJ3bxRxxZvnutipePf4KAB/rupv5rRdR/tUbBogrgrMXwzA43HMYCQmf3UcgHkDVVeZ55xFRI7SET9AR7cRjc7O/az8+ezazPXNYmGMKarqmE4qG6NA6KHQWpoWT7hNBEhEzRTRndjYWuyXjPUHU6DxfSepJdF2jPdpOTIvhUBwUZBWknY2nw0Se28/XeeLjv7gz/W9LwopqO5Wi/ezdv56KkNANnaM9R8yun4aBJ5KN1+XF6cuakngmmlgslr63djgGb/o1Vk6808ovvz72VL27vl0+aS41wzBI6GaqriIpaIaGYRjYFTu6oZPUk6i6iiRJJLQEiqRgt9ixK6YL0dANdENHNVQssimeGIaBmtDMpgcSGXNE6j3FHHF+kro/VXUV3dCRJRmLbP7+ZWng9eNov2djObePqynBcGIamAXpzmZBTSAQDE9/d1pOSwHX/HYFEhJzP1iIIevIukJWs4uDl31Ad0Fma2VPVzZd73TzTuO7ZBXb2Zqo5VjoKI+/+yP+qtR80qmpGjaXjUQ4wezSGRnbhzrCPPX5Zym6bBoXryhhwY3zaH7zCDtr3yVwrBvfrGyWrr6c4mvmnJHjIZhYPvu7B9L/XlywmN3tu3nw5S8DU3fBLpg67ir5OG+2vAHAwhXzuXX2jeIC+Byjb0FxgGx7dvpGpq97rSPSgaprWGU7uQ4z88FMAdLo0rs43H0YRVIolKYjy4p5hySBrMgDbpTioQThjgjWLAtZOVlY7RbioQSRrihaUkOxKjhzsgakjQnODQ4HDwG9N8RIdEY7ORQ8hM/u4/J80cBksniw7CH+ZfujzN97CSXvXMobd9YT8Yb4SunfTFlMsiTjtWXTGevAIlvJmZ6N3TIxwtP5yrYnd2d2cR0NEpPqUpMkCZtsS88TFizpeUKRTrnXNF3DwFxukczzvm7oZkqooRLX4iCBZEhpIc0wGLREhK7qGLqBJEvpjBldM7tOS71ziyg9c26S0E7VUVQkhbgWJ67FsUgKWWfIzTwmQe3FF18c9dhAIEB1dTWf/exnxxyUQCA4O3jyvZ8gS3Ja/b+k8cr0OgkJSe/tyCOb67Pbc5l5YD6HLvuApD3BzKa5IMHXjlTCkVP7bQ42nRJO7vs1S1ddTkezn+wiT8b7H9vVgpbQOLarhWkX59P85hHqH3mFWUuKWP6pxTS/cZT6R15h5YYbhah2DvJg2UP8y45HAdjTfqoQ7vSs6RwI7J/y1BLBmeWS3Ev5cumDTHNO45LcSwesj/XEefvxRq5+oEyII2cx/UXQ1GvDMBN85nrn4rF6aAm3EFZDuCxmx76D3c2EkxFiapSw2sPu9l0csvsochVyUdHFGLqBlhzYGTQZSaJrZr09h9dBPJQg2NqDzWnF5jr12lvoEX835yDTnYWcjLTSHe/GKltpChwgy5pFQVZBRg0+wcRy06ybeeWZN5i+zaxxubz+Jrbf/QrXzrhuSuPKceQABh6bN+1C6UsyrpIIJXDlibTgZEyl7cOOsYlpAAa0fdBOMq5itY/LezMkw80TADbFhoycrrWWcholtAS6oaMZGrqhE01GScpJLLIFh81hbj/I5zV0A8MAQzOQZAMtqZvimiJh6Hp6bhGi2rlHqiafgYFmaCT1BAZgUZT0/etgTrWJZEzfkpGaEPTndJsSCASCqae/O01JKvja85H6t4MCZF3h2ufKkXQZCYl5759yrwEsffUadt4wMCX0wbKHzO0VmWkL82kJt+CyOGnubqZmTzV3B1bjzncS6ogwe+kMXt+8jVlLivjYN28lHkrw5g8bkWSJV//n7bSgJhxs5w43zbqZXzf/ig+63k9X0pCQOBk9KZxqFyg3zrppwLJdbTt5bPtmrv7dCsIHo+x/5RBgkDPbJ77f5xD93Wu5WbkZaTiphgZdsS5ssp1QMoRdsVHkMp3LkixhsZuNDqyylZgaM5uraF4k2bxYtmZZCBwLYnNayZ7hNZ1qiSgRfwQtqVN4aQGAcLCdQ3hsHrrjAQ4GD6LpGgk9gW5oJPUkR0NHAVjgWzjFUZ5/qIZK8BI/vt3TsPuzaFtwnOk505DlqRccchyZxdEjyQitkVYKLAUk2lRioThdR7uxe2xYbJYL9vttdVj40x/dm06XHws2p3XSxLTh6O9esyrWjHkiJZ4k9SSSJPWKbVJaVJd63cy6oSMhpVNIFRRAQpIldM0U01LlZLSkgaZq6KqBzWlNjxEOtnMDRTbThmNqDNVIggGypJDQk6CCJMk4JtnJOqZvykhNCPojmhIIBOcuT773k4zOnppV44X7f4ElMfSTYAmJEt8Cvrq8MmO5NcvC13c9zIHA/vSybFs2N826OWNc9e7v0di2A5tsJ6HH+U3es3zh0S8id1iZVpxH4Fg3yz+1GEmSaD9gdnQxdCPdNr2vg61gQR6h9jD1j7xCeeUNlPR2EhSC29mDaqgoskJhViGt0VbAvFhK1db41CV/SlukTbjULlBSNfZq9lQTaOsmeLwHBQuGrnPlJy6lo6lLOFTPMYZyJcCphgam80BDM3SybT48tkzn8pHgYVRdJZyMEFEjzPXOZXrBNAxNQpIktKSG3WPWVFLjKgCyVUHXzSfVKcea1WFBUzXUuEosGCN3bk76plsIbmcPpmNF4uKci2mLtNETD5KblUcg3k2uJKcFV8HEYpWtbLztEaJLY7R90MFnr/8kOvpZ5QbsW1+vOdCMbtOxBG1E/XHc2S6cOVmoMfWCdqhmZTvIyj630mKHmydSDQ1MTNez0q8IvW7oxLU4EqAbpmtJskhYJfNvV0sayMqpfRqG0SuiGaYYp+lpB5skg57U0VUVxXaqYY4Q3M4eTrkarci6bHaH7a2rpxkadsWa4XKcDMYkqKW6WozEiy++SG5urhDUBIJzlJga44OuD/r03zFJOOIkHMN33dtr7MFRYM2oa5HUzadjC3wLuW3u7Tx/6LcYGKiGemqCMzQa23aY76Ob79HUfYC/+f1XAPgn30akAoP3/3CAS/5oAVaHlZxSD527A2T5zPfaWfsus5YUccc3bubxT9WixlQUm8Kuur2UXDs3LbhNuzifi1cU03kwMOgN+fksuqlxdUANookcP1paw608f+i3RNUof7zwXr63578B0mIawJPv/5Qn3/+pcKldoPStsYcPIs4QWbqLNz7WwNo//xSGYfDcN19kZ+27583380ImlRJ6Uc5F+Ow+/DF/ennqhiquxjjac4SoGkOWJAqc02gJn8Af8+OyOimUi4gSxegxcPqySEhxWrVWbFE7eT6zG1ukK4rNacWZ6yR5TAUZEtEkka4odrctLbhJsoTVZUVPaIPekAvRbfKJqlFCiR6SehKrbMVpyUKRFVxWFz3JHgKxLmyKrTcNUDDRWGUr1ulWvNNNUdt0+ZgktSRvtLw+4MHomeRgdzOhRJiO3oYVB2nGEXVhkS1cvLgkLXB0nwimv9+Cc5uUeOKwOMyGBroGUuY8oRs6SS2Jjo6MjCIrJLUkmqQhSzKSIWFopqgmSRKGBAktiUVSzIcympZ2sGmqKazpqo6e1FAsch/BzRTTMBgyZVQIb5NPKq1TkmRARZZlDMPoFdV0NF1DQ5tUl9qY7pJG22CgrKwMv9/PM888w5IlS04nLoFAMIU4LA4ev+MnhJPhMW/rsroGFIm1ylaqbvwuFsmCJEncPu8OU0zr86QzrsZZNn05208ObCf9YNlDvHzsJd5ZuI+yl25gy98+y4yl0zjWcQyXls2Bq/ayr3MvXccCfORTS+g5GUaNmU+wsos8BI6Z6ecpwc3mtLL3OdMtV7RoWsYN+Z5fvs+bj20nb14OV959KW0fdJ5R0S2uxdOdjCZ6/Hu/28+urfu461vluAtcI44PtYf55TcaWHLPZVx62/hSavp3WOpJBHlmfy0AjW07mOmeyfHQcQDmeedxqLcQdSotWHDh8fnFX+R7u/87/Tor5ObDJXtYe5tZm1WSJGaXFvGHn+4iEUlic549zgnB2BmuocGpMTLzsos5GjxMUldxWpwE4t24rE6KXDPwxzqJ2KOcbDuJYWgEpG5OhluZ5pqOJcesv5NysKkJNb1fR7Y9fUOUEtzUuEaiJ4FiU7A5rRk35D1tYbpPdGP32Id0wQjBbfzE1BjdCXP+dls9dMW6sMs25nrnsb/rA0BiWtb0qQ3yAqP5zSO8/eOddLd205PdzZE/Os6frfrUlMRS5JrBQbWZUDKE2+qmJ9GDx5pNcXFxhmhhdVoJtY/9elZw9jGgoYFsGThPIGFTbCT0BBISsiSjGVq6+2NSTqKrGlpcxapYUXvFNovNFOgM3UBOdRdPdZHu0+gglTKqWGSzsyhg6Obyvn93akJDVzWkXhHNrAPaW9NLEU63iSJVTw/MpgQYGrKsoEgyyd7f32Q7a8dtOzh06BCNjY34/Znd/QKBANu2DbwxFggE5wbZ9myy7dkTtr++JzNJOmW9TuG0OvnG1X/P3/z+KxmpoQt8C7lp1s3UfVhL67xj7Lz1debvvoTu2hDJbJXtt77KyenH2PPqDq53386eN/dx+V2X8Kkf/DEdTX4at7yLvdDKFxo+x9KjN3DVny5l32/N/csWmbkfmcn2n50qiL/nF2bduM5DXSy68yKWfXJx2gUz/6rZvPVEI/Fwgg/qm9LNEY42tgwQ3U5HcHv+0G/Zuv8Zvn3ddyhwFox4TNsj7Xz99b/lnoX3cvu8O4Ydq8ZVdm3dR7Clh19+vZ67vr1yWFEt1B7ml1+vJ9gaYtfWfSy8af6YnWqHg4d4/tBv2dnWyJ9f9hdcM+Pa9LpiXwkem5eeRJCIGuHfbv5Pvv76w3zQ9UFaTLs455IpffotmFrumPdRnm36P46HjgEQzg4yt72EG2eaddYMw+BoYwsWu5VnvvJrbvnKdRReMvL3RnD2MlyqD5iFqouzi1EkhUPdhwglQqi6il1x4LQ4Oam3YnVZCLt7ONDVTCwWw2630+MOsLOnHYI6hYmZ5PcUkD8rF6vDihpXifgjJKQEzd3N2CIOfHnZ6bpDsiJjdVqJ+KPpOALHg6gxFbvbjsNjR8l1pl0wik0h2NpDxB/FmePAmZtFMpIUgtsQhJNhehI9RNUoczxzMlK3nFYnnTHzBtlpdTLbM5u4Hqc7EWCaazp22Y7X7p3C6C8sUi5/AAkZT5ePyE8Mdua/w9Kbz0zH1dZwC8/sr6PQVQSYD2NzHLlE1QhOiwuP04tdzXzImAgliHTF6D4R5I3oa5yMtfIXi/5ywL6/8tJfs+qiNSwuWALA84d/C8C9C1dN7ocSjImR5glFVlBkBUmVzJqLfbqFSr111GQLqKqKqmmAgWKVUVGJJ+OggR07isWBbJGRFNAS5vaarpFQE9istrRbDkBSJIx+zRC0hLlvWSGdKqomNHTNFH90TUdXdWSLENxGQtXVtGhmUzLnyVTzPAmwKva0wGZgCq6yJGfMK5PBuAS1nTt3UlZWlm5SkJtrFon0+/2UlJRQW1s77gAFAsHoCLb2sGvrXryFpjXf5rJx2e2jdxXte34/wdYerv506YB1zzz4HEvvXcTMxeYFzHu/MwWpJfcsmoDIT5Gqi5BKDf3d4efTy79S9iAf+N/Hf1EXs+6Zle4O2ZcDV+4l+6VcfvMPLzG7tIimbYdpP9DJrhVvcjx0lAXZ3bz/9gHmrpyB0SWhJjSO7WzFN+uUcBjpOnXT9PjRH/DJ9vuYXVrEtid3E/ZH2POL9wBweO187Ju3IkkSV3z8Up775ou88dh2gi09RINx9vzvvmEFtxR9hbegN4BxuczXjA380/WPDCuqtUfa+dprG2iNtLJ1/zPcPPuWYZ1qFruFu75VnhbJhhPV+opp3kI3d32rfEQxLXVh0ffCpi3Sxq+afwnAzrbGDEFNkRT+9iNfY4Z7JjmOHLNorGzh4pxLWDGnnBeONKRrZfQXXwUXBqqhkmXJYo5nLh+d9zHebt/OzF8t4NfffIG5ZTM52tjCsV0tAMR74vzmH1/kU4/9MTbnhSdKXEikUkPnZc/DZ/cRiAfMmyUkZrhmEtdiOItchJI9dEQ78dmzCcS7sRkWPDYvKAZ6RCfY0mMKZaEowXCQFvsJ/B2dzErOxtqj4MjJQtYVLFaFRCSBYj11Qa7Gk1gcFmJqjEOhQxS5i9KimxpX6TrajSxLOOb5cPqywJeVFtzUuEoyphLpiuLKzcLuGVxw68v5LL5F1SihZA8AETWSUTPPrtiZ4ZqJw+JAQiKqRnEojvTDGJAGuFMEk8eOp/ZkvJYwC8A3/e+RMyKotYZb+MrLX2Lzyh/itrkxDIP/2PnvRNUIHy+5m4gaIRFOEg8l6D4RxOq0kowkaTrazAvRepT9MtuCf+CO4sEfQDZ1N1G1bWP69e3z7uCLS/7fpH+u/hj9lRnBmDEMsx6aTbGZqaFGb0qmJGOVrRiyjkWxoBlauuaWpmtISCgWBV0zUBNm6qehmwVwDFkjqam9xe4l7DYbkmK2atM1g74lugzDMGOQIaEnkXTJFPl6a7UZeu/+AdmiIPc64NSEhqbq5vY6GJre63CTBhXc+nI+i2+6oaH3fi/6n/MVSUFW5HTTvJQbMSW09d9mMr5f4xLUampqaGpqYv78+ezcuRM4lRZ68OBBAoHAuAMUCAQjE2zt4ZkHf8Of1Nydvsh+64lGdm3dO6zolRLhAJpeO8Klty8YdFxHk5/6Ta+mX196+wJu/MLVE/gJTIZLDS3xLaDEZ8ZnGAbPNv1fhpPNrtjxF7dRtvhyDv/yBNue3E2Hu40Dt+7j5BzT4fL+5btxveTlw64P+Pgdd6ZvyJs++i67Xn+Nf7z2n3AXZqFjcGjOfpodB/jhnh9w/du34Z7houfkqZQB3yxv+uScSj07truFt58wz4UzFxdmCG5b/uqXvPgvr/POgve5+i+WMv3igvTT3hlXTGfWkiL8LVk4X/Kwg1f5GoOLas1vHmHb07voPNrFguzF+Jbn89VPPzSqtE93gYu7vr1yWFFtgJg2iOi2q20nP3r3Bzxw+WdwWLKoP/w8jSd38NXlG7gs77L0uCvyr8QiW9ANnYgaGRDPovzLM37337run4ZNCxZcWPQ/H3ys+E4OXHaIPc+8xx+e3E1WkZ3rP7ec/S8d5OQHHVzzQJkQ0y4AhksNdcpOnFYnOcDh3uYFKRebx+rGZ/eh2wx82dlEuqJE/FGORo/Qbe2my+hEkRQ6rR2EOyK4Yy4uKrqIWDjG4bbDeArdaNEk+Vn52HPs6JpK2BnhRM8xZFnGHXSDIqMlNXRVx+a1o1hOiXBWp5VwRwQ1rhI8GUp3IgXSglt7UyeuXCeSLJFd5EGSpXRNN9kiI1tk1Liadj4PJaqdDQJcJBmhI9pBflY+mqERSoaIqTHmeedl3AyZabtdSMjpm96+OK3O9L9numcOmxIsmFy6jwcHLjQgcLz7jLz/M/vruH3eHbhtbsA8F/zl5Z/hU8/dz2euWGuG4zFIhM2aiP62AA67jYsXXMS88Dx0Tefo4SND7v/2eXekrzOXFCxJu+DOFFarFUmSCIfDZGVlndH3Pt8YkBrKqdRQRVKgtx6grukYGGkXmyIr5jWHLIFhmEKZBJqsktST5jlKAVVLosc1FEXBKtl6HVE6uqah9F6zyBYJVVdRUZF10x2n6zqS1JtJapjOtr6nMEmWzCYIBmiqhqLI6W6kYApuyZiKYpXNz2I99TlSYpskmymomj54XbcUZ4MAp+kaqqFikSzm72FIF5qCbqjpZnkp8QxIOw9TWGVrxrzQf56IRMz7Eat14u4vxiWolZaWMn/+fACKi4vZsGED3/ve9wCYP38+L7744vgjFAgEI7Jr614uvX1BxgXz0nsv5/FPbRlWUPMWetLCWPsB/5DjLr19AfklZkHnWYsL0y64yWCk1FAY2sn2nRsewS7bWXbLEgBePvoSr+841Zm4dd4xdtzyKtceuJVtT+7GNyub925vpLnwA/JD+UiSxMsXPU/ZSzeAoTBDnUPBa0V0nQiy/dZX+eP5d/BH3yrn5f94EzWhsqdtN9Nc08h15HFkx4mMGOeUzcgQ3Jy5WQSOBWnd14YaNye5VE23az5TRu3/+xUA9mwbl7y7mJfn/ZqvvWaKau893mTeiCU0Trxzku7Zfo4vPcTM1nmU/PYKjuS1QqmMt8iD3TX8TVN/UW3rQ78hy+sgeLIH17QsgvEejDYpQ0x7t+Nd9nW+S0e0g7Lpy6j98GkOBg/y0/d+TOm0MuoPm8e48eSODEHNYXHwrWv/iXnZ83FZR67ZNprfveDCov/fxMLr5lO0PJ+HX6ukJdzKRUs38PHbbqP5jSOUXD83Y9vBXJP9maymG4LJZaSUn5SLba53boaLbbZnTnp8ar60J6y859+Hv8fAYXFgsVnIcWWTncgh4o+iKzr2aTaMLB1VN+eeLlcHXS3daIkkUTnC9uM7sCVtlBTP51LvIrxFHjAMDKtOTI1hkS0kwgmQIJaM0RpsYXbe7IyYrU4riUgSu6u3tlvvR0rVdLO5bOk6UJJMRk03wzAIHA8iK2bBbDWmYnNa0+43/5EAnmku7B47FpsyZhEqJdAlE0lki0RWbhYujzNjTE9v84CEFifHnoM/1klCj+OPdWIYBmE1jEW2EFWjGSJZliWLGa6ZZFmyRoxrpN+7YHLJnunFfzhARr8qiQyX/4nQCWa4J6fz6mvHX+XTix7IWJZyNO5q28mSaUtPfbcdOv5QB5ocJ9vlNVOvYyrS0Hoaha4i7pj30UmJfTQoikJ2djbt7e3E43G8Xi8Wi0X8nU8ShmGkm6YZkul21dCQ5d50S0nvHQi6rpNUk2Z3+l7nk6zK6JKOJmkYspFO5bTICqquoekaetJ0rkVjUdAlFGScDieSZIptaKBK5jlfQkJLmqKchoakSVhsFjTp1IMGszO1hqzKSBJYNPP6xawJaopzup669gESYLGZY3TdMGu6pZo36L1psL0uPF3XUXpFtb4140ZDX3EOCWRFQpblAcJW6qGJ2Z1VQdXVtOCVmrclJDRFy9g2JbTJkkxCTXA6GIZBJBKhra0Nn8+HokxcGui4riL7ftDs7Gy2bdvG4cOHmTvXvKhtbGzk1ltvHV+EAoFgRJpeO8JVn85sGpK60D62q4VZS8b3lM1b6BlT+uhkM5omBwA3zbp5gJPNvSSLT39pjVkLwdB4+aVf4Yq4yHPkA/An997Pz3iKkj2XcdHOKwlld7P91le57Y5bsTltzLyykGs+U0b9I6/w3t/vo31mCxd1Xoal2cHJ649wXDnK5TvLeO3lN3l5Vj2zvLP44wX3EGwNpWPolgJ0d/rxHw1Q9snLSUZPFcfOneVDPaBR6CykNdLK117bwDW7VpI4qYIE3bP9vLbieQpdhfzFdffzwoNvsP3J3Wx/cjdXf3UJ8sWmG+zy/CvQuwye+vyzWGwK06/K573rG4mqEZYXXsVd317J0198lmggRiwYp3nZe7iPZVPQVkSWz5HhTNvV1siWD58G4LeHfgPAPQvvZev+Z/iw60MAbLKNmHYqXTZFXxeaQDAR1O2v5XDwMAA/fPcxlt5ayoIb5g0Yt/e5Dznxzklu/MJVOLwDHZwn9p7kuW++yJ3fXEHRomnjiql/XR+X1TWmG7PfHvoNreGWQev6CMbOaBocpPDYPEx3FhJNxtI3Qm6vm1leUxSIqzGikRCqrmGVzcvmOdPmAEc4ebIdwgqGYeCa7mTutHkoFhnfDC/B1h7aTnTQIwfRohou3U1+US7BZJyeth788U7kHh2LbMFtdaNFDLPOjmTGH9fMbtfRWBRfQTaGfkrFsLls6QczYKYbhUIhOmOduBJups3Kz3C/BU+GOPlBB97pbjxzXeiYrgyX1UUsGKenPWTeWGVD0hFHNwxyHbk4LA7aD3QS7oyguBSCngBaVMN2yM78+fMyHuJ1x7uJaVGO9hxllnsWkiSRY8+hK95FOBnGH/MzzztvgAtNkqQMgU1w9lJ2/5VmDbXezoap/5fddwVqQuPZHz/HT3Mf49NXPsAnFtw9oe8dSoQIJ8MUOgsHrHNZXTR1N7Fk2qnr4PZoOwY6SV3HH/Mz3TV90AeOwdYeFJvp4ky9z4HAfjw2T9qtdiYpLCwkKyuLtrY2gsFBHIGCKUPVVVRdTf/5W2QLlt45QTf0Xpea0euAM1MPNU1DMkzxTJZkLIoFq8W8VzF0U4TT0dOdSmVkDMkUnmRdNkUpxUytlpExNCPdFEGSJZBNIU5XzdRQDCM9V0iKhKGfqt9m6AaaZtYXk5EHONLUuIoBKIr5nilxS5KkU2IZgJz6h4EiW9K14EybW6/LzADFogyoXZYSMHVDQ+49TqnUTN0wU10tsgVFGvuDn9Hi8/koLBx4HhkP4xLUDMNgw4YNvPDCC2zbto0NGzZQXl5OTU0NXV1doimBQHAGiIcSJMIJvNPdA9bZXDY6mv3jFtRS79N+oBO7x0ZBr1ttKhmPky1Vl0uRFP7z1v8BzBbw0CvCLfk/Xp93ytlmla3cMe9j6dfF18xhzrppBJ7pTItuB27dxsm5ZmqpIeuUvXQD/sf87Cp+F3vIR6gtzG0bbmTOspn8YO9j/OrVZ7neczsHth/i0hULufUr1xEJRNn30gd0e/3oho7T4qQ10kqoO4wNO4ZhcLzwELIs47F5meaahmeaKy3Wvdn+Bs91/B8A/3zTv5Ifn46u6iRUnWQiye+PvQxAftY0bp93R3pyNHSDI0VNhC4LctXzt3CR7ZKMNM+sQW52bp51K1v3PwPAmovuZ/XFa8bUnVQgOF0+demf0RHtYHf7br5x9TfTF7R96ToS4K3HG9ESGm0fdrD6P/5oQMrbtp/sQkto/OGnu/jExttOO57+dX0AHt/7Q57ZXzdsMeuUCAem82KkxiKCsTFaN9NQNdlSApzd4mCud156LJginNvrppNOLKqETbXgc2en3TJ2tw1voYdgS5BEJIFslQm6AwSTAeJaDN9sD22tbRzadxCny8GCrIvwGF4KL52G3W3D0A2Oho6S0ONEkjFcESeufFfagdbR2UF3Ikgk2EO23YdFtdAV7+JkpI3sRJxEIkZ7wIrPnkN+Vj4Wu0IsaAp0bdE2EnocCZkSX4n5mQzzs6m6Sihpzide3YOh2wmeDCHLEq7sLOKeMGRb0TukDIccmDdFAHmOPHRDR5EU3DYPXfEusixZXJ53BflZ+cJtcw5TfM0cVm64kcan3iFwvBvfzGzK7r+CeVfN5tnvPE/7tm7Kim7gCfVxknqSF4+8wLorKzKErtPlZKR1yHUeq6e3pt4pCl1FnAidQJbkIevRxoJx4iHT7aLGNXa17aTQWcjigiWcjLTyjde/xl8seuCMCmuSJOHz+cjOzkbTNFRVHXkjwaRjGAYnwmYmitfmJdj79zbDNWOACwvM3+Px0HFOhE5gkczzc54jjyJvETmOnPT4RDjBkdZjJJJxOhMd+PJy0KxJ080bSRDpjGNYNQpzinDhxqpacee7zIcqmsrxsHnfIXVK5LsKsHvsaYGrvbuDeDKOs8BGXlY+8WCclpOttEc7cRku3NOdWDwK+Y4CnFYnre+1Ew8lyJ7tIejqAiBLcTLNNY1Qe5hExPyuJHLixIkBMMs9m67DQWI9UZy5WeAy6LEFiXUkyFKczJyb6VY9EjyCgY6qaVh6HWJFrhm0hE+g6Rq5jjx8Dt9E//rSWK3WCXWmpRiXoLZ27Vo2b95MSUkJAKtWraK5uZkVK1YgSRL19fUTEqRAIBia4MmeIdc5PDZiPfFxv8exXS14p7uZubiI4MkefvV3DVz16aVnhbA2HKN1sgFYFXPZUCKcw+rIGH/RdSW0zD1CW+QE052FvH7gWHpdKrW0ZM9l5L01jdjcOLdtuJH5vQ0JYrrp5Eo1UXjxX94wa7DtbCF4KMz7t+6hLdqW3l/9/Vuxxe1c/ZtbKThexMFFH6TTaGKhBFnZDuZfO4fm/Peg09wmokaQZIn8klzUuIo7/5RAFu2tZ6YlTzkF8k5Oxxny4FhgI/iHzL+pa4quYZZ7Fk6Lkx+++wOaug/w1y99ETA7cX7q0j8VN0mCM4ZVtvJg2UN0RDuY5hzcWRb2R7HYFbSExryrZw8Q006800rre+0AtO5r48Q7rcy44vSeWPav6wOwauEa/uS5+4YV1ApdRemC1we69g85TjC5jMXN1reo8WAppX23s7ttTJ9bgFd1kdTNBhvt0TaiahSf10s4EcHRbccTzwaHlNGQQJIlDMwUF4fPbnYc7QibRdajSZIJFUu2TEJPcCh4EN0wiBVEyc330n0sSOvJE3hlF+48U+CTFRlnThaObAdxyZx/DEw3gCRLWGyKWUfHcsoFl3ILpOrByYpClsVpCmdu0CKZTrMc819XhwAAZVNJREFURw7ZRjayJNMRaSeuxznaY+bXOS3OUXWuFpz9FF8zZ0Bzpc5DXXTsNm/Afe15fMRxDa8de5VjoaM8sfdx3DY3Xlv2kOfriSCcDGe8tsgWZnpmAqfE3v7oup522Tk8Nr645K/SLmO3bQF3zPsoVX/YSM1tP5i0uIdCkiQsFgsWiyhJcLZQ7ChOn98LKBi2hqNhGCgJhTm5czLmiEJPYcY2DoeDma4iknoSX8KLaqh0RDtx2p0k5CRuixN6JNSQitVrJW9mXnqeSOpJlKQpDjnz7BCUMKIGNqeNZCKJmtCw5iroFp2W+AmiSoxYQRSv4aLzmJ+THa14rS6Ksotw2Bw4HHZsFhvZPi9hyRQMZUXG4XCgZmkohjlP4DBQe51mNrsNWZOxOxxYFRsWhwIOA7svidxjweHIvHeaIc8wu25KMu3RdmJajLbkSRSbgktxUeiZWOfYmWLc39K1a9dmvF6/fj3r168f724FAsEEkQifXq55X278wlXpumkF7jwuu30hDZte5ZPVd49735PNWOtyjVaEK/GVUOIzHyYYhsE7HXsyUkutV8h8/M9WYpEtLMy5KGPb0ulluG1uYvNiXFdaxge/OJiu6XbJF+exzXgBj+blyvwref3EayBBwhHnw6XvUPbSDVxdfyuJ4ijPvfkinc1dabHOdVLB6NBwWlxMc07HV+Dl3n8xnXVJPcmlkc04rU5cFlNc883MJniyBz2pc/lbywBQbIpZ/6cPM9wzmeGeSVJPYrfYRSdOwZQjS/KAmzPDMHjp6IvcMOtGZi0pYvW//xGNW97h6k8PdEdse3J3um6IJEts+9kePrHx9C7kBqvrkxLXUnV9BGc3Y63NNVoRzm1zZwitwUQw3SjB4lSYM62EWZ5ZWCRL+qFOCq/Ni6pryHYJt9eTbqKgWBXc011ELRK6oTHDNZOuuJ+oGsXlcBHOiWJrs+PpySFhSdLdbdZWyy82mxjocZUs3ZEWGRweOw6P6S7WDM0UxTA7pEmSRO5cH5Is4Znmxmcx00i7u4Po1sxOaVkWM2XOPA6y6MZ5AZE3L4c/+nY5z2/8Pb9f9jLtegv0Gsaaug/w4MtfBuDZu3992u8xXB3WnuTgD5ZTqXd9MYCEZl4XO31ZWB1WEuEEdredQjKzOUp8JbRGWsV5XACMbZ4Yy4OalGNtmnPagGY6s3JnMXPODDRDx67YM1IoZWRy7DnohoE9y4bdmZUxT3inO1HtGoahM9M9i5bwCaJqFLfDTbcvCB0S3p4coo44htqDxWbBO8d8sOOL5iAhpeelvg/lVd2NgdErjCk4vHYkySwPJMkSOYqP7mAQ3Tmwm2ZqPjR6i62dL/PEuAS1rVu3cs8990xULAKB4DQYrgh9rGf8YhowoAlBfkkuwdbQhNRnOxsZqwg3lKttQc7CQd1w18+8getn3pB+ffmNl2asv4nrAWgLt/F2y1vp/bfOO8bOW9/gloN3EPpDmNisTOdb6fQySqeXDfmZ+hYKDrWHiYfi6EkdxaZwycoFvF9/AC2hEe2OEWoPD+juKTpxCs5mtnz4NE++9xN+f+xlNnzkb3HlObnh81cNGPfOs++l3WlgpjyfrkttrHV9BOcPYxXhhnK1ORTHEDdZuRmvh+vSGVEj6Zsw2SlxUclC8tR8tJiGbjUy3G/Z9uwh96NIygARwp3vItjaQ6gtZDrkIkkSkeSQzYkkSRLdOC9Apl9cwCer72ZOWx7/suPRAevXXfG5ce3fbTX/3vo70VLLRtP4yExrThJOhumMdpKXlYfVYcHqsPD43h9yw8wb0+md3SeCxIKmC6f52EEudS46451yBec2EzVH2BT7oNsqskJeVv6pBfbMeSKHzHO9P+ZIzxN2t5352TPNeSI0cJ7I77vffvQvs+HKdZpzRHt4VHNE6licT/PEuAS1yspKysrK0k0IBALBmcfuNp8sD+ZES4QT2Ebo+jgSbz3RSMn1czPSO1PvGTwZGmqzC4qxpJaOlvZIO197bUO6nfRdJR/nl03PcmLuYRou+QXfuaHqtFNoQu1hfvn1eqKBGFk+B1nZDj5oOEB2kYdod4xoIMYvv16f0Zig72dNITpxCs4W2iPt1H64BYCdbY3sad+NXbFTs6c6o4ZPPJTgzccbB93Hq9//A/f998fH9L5jresjuHAZi2NhLAx1E+bzeCfk5iRVDy7D+dDnxmswRDfOCxOr3TKgGdSs/fMxrDpP259itnc2iwuWnNa+3TY3LqtrSDfakoKRH1yEk6F0t8DueACv3UtSS9IaaWXr/mcodBVR4ltAqCPCyQ876FbMNNbpWdMJtvaM+HcvEIyHyZojYHLnidOZI+D8mifGJah1dnby/e9/n7y8PEpLS0VHT4FgCrC7bdhcNmKhwd1osxaPz0G2e+s+vIWeDEEtHjLrsg3WCOFCZSKFppSYdjJ6kulZ0/mn6x9hmmsad86/K738a69t4J+uf2TMolpKTAu2hvAWugeIZn3XDyWqCQRnGwXOAv7x2m/zrbe+ycfm/xF5WXn8987/GlDD59AvTmBoA9MQAALHguOqpTYYg7kpBBcuk3EDMZk3YSnsbpsQEgSjoq9j/4bwLZx8vRsMiX2RHWzxPMWV+YtP+2/z+pk30BpuyViWej0aJ7Db5sEiW5CQyHcWoOsa7dF2jvUc4xMld3PLrFtI6kkCx7ux2i0cVg7isri47tLr6D4RHNCIQyCYaCZLZJrseeJCnyMGr9I4Smpra9m4cSMPPfQQOTk5fPe73+XRRx/l0KFDExSeQCAYDSXXzyHYmvnULvV6vCmZV316KZfdvjBj2bHdrdhctvMy3XOqSYlprZFWCp2FfOeGKqa5zFpR01zT+M4NVRQ6C2mNtPK11zbQHmkfYY+nGElMA3AXuLjr2yvxFrrTolqoXYgCgrOfy/Iu439WfJ/aD5/mwZe/TFP3AeBUDZ/P/u4BWt49Oew+3v7xrjG95+nU9REIJprz6Um/4Nwm5dj/55v+lfyjRUiGjISEL5LHfdqfUfflX/PYqp9R+6Vf0fzmkTHt+96Fq3jj+OsZy3576Dd8YclfpV+HEiG+8frXaAocGHQfMTWGIltoDbewz/8eLeETOK1ZZFmc1H5Yy8HuZhweO5ZpMr/t/DVfXGo2jrE6rWhJDUM3MPTBH8oIBGczYp6YPMYlqK1YsSL976VLl/LVr36Vhx56iIaGBm6//XYee+yxcQcoEAhGZsk9izj4RuaFyb7n93PDF07VD4qHEvzq7xpob+ocdB+JcGLQtNH84lyaXz+csZ/dW/dy0xcH1iYSjI/+YtpgDrQCZwH/dP0jYxbV1LjKL7/RMKyYlmKAqPaNBtS4aN0uOPvJceTyYNlDg677mO+ujNppg9H2YQe//aeX6BmliDwRdX0EAoHgfMIqW5EkiesrllO65nLmXzuHT9x6F2/+y078hwNoSR3/4QD1j7wyJlGt0FVE5Uce5vG9P+T146/xzP46PDYvd8z7aHpMKNnDgcB+ehKnHmiEEiEe3/tD/nvXf9IaaeWNE6/xXPOv+UPLW+iGgdPiJNvuQ5ZkXjzyInUnt/D4sR/yV2V/na53m4wkUawK0e4Y/sMBIl1RIawJBAJgnCmfwWAQr9ebfr1161aqq6upr6+nuLh43MEJBILR4S30UL7+Bt56opFpC/IIngzh8NgznGXxUJz2A37ifRoVxEMJdj7zLolwgmBriFhvi3tvoYcl9ywCTIfbsV0tvPWEWXeopzXEDZ+/SrjTJpi4Fufrr//tsGJaipSolhLfvv763/Ift/4XdsU+5P4tdgtL7rmMXVv3cde3ykdM40yJar/8RgNL7rkMi120bhecG/Sv4ZPiua5fckv2XTi7h09VP/yH48xZdmKAM3cwJqKuj0AgEJyPSJLE8k8tQdd0nnnwOZAw22xi/t/A4KUfv8b8qz85ardMiW9BunHAYBS6ivj5nVsylrltbv5i0V8C8MUl/y+9/HDwMIeDh+mMdhJOhpntmcNsz2wK5SJ6ToawqVYigWi6yLpnmptwZwRd0wl3RrC5bFhsAzuJCgSCC4tx3SGtXr2ayspKamtrqampITs7mzVr1rBjxw6WLhUXkQLBmaSgJC+jzll/vIUeHvjZmoxldreNqz9dCsCNX7h6yG1nLSkSAtokY1fs3LPwXrbuf4ZvX/edEWujpUS1r7/+t9yz8N5hxbQUl962kIU3zR+1OOYucLH63+8UYprgnKJvDZ9l05ezdX8dCT3BjOa5I4ppKbzTRu8sG29dH4FAIDifkRWZ7uPBU2JaLxIS8dYk1Xu+z+cWf37YfcS1+Kiuc0ZL3yLtVtmKLClohkaO3YfDYXZV7F9k3eKwYI1aiffEsXvsQkwTCATAOAW1+vp6GhoaWLFiBVu2bOHee++dqLgEAoHgguP2eXdw8+xbRn3RWOAsGNGZ1p+ximNCTBOca/Tvurvm4vt49N/+laxG34CxbTNPkH+iENnoUwFDgh1b3mXW0hnpRbu27iXY2sPSVZfjmZYpyt27cBV/9/rX0w4IGLyuT9W2jfzFogcGdVeEk2HRwEAgEJy3ZM/04j8cGCCqJRwxbpt147DbvtvxLt988xv8wzXfYlH+5RMST/8i7dNd0wnGuvHas4FTRdZ1QyemxrBbzYLr3ulu1JwsBjPUBVt7sNgtOLx2ZGVcVZUEAsE5xLjulEpLS3nhhRfIzs6eqHgEAoHggmasT2An8omtQHC+0LfrbtveDnJ+P9Bh65/exrbbfo+7y8uit8vIb+nt7mlA6762dMfPRCTBrmf2Eg8l2P/SQT71g3tweE997/rW9Vnou4jWSOuo6/rU7d9COBmmNdJKz/FX0/u7d+GqiT4kAoFAMGWU3X8l9Y+8MmD5jHvyuSxv0bDb/nTfEyS0BD9578c8csOmCYupb5qpIinkZOUOGBOIB/DHOnFaXBRkFWBVrIM605LRJPFQwvwJJ8iZJe6NBYILhXEJag8//LAQ0wQCgUAgEJwWwdYedm3di7fQLO5vc9lGVbssxb7n9xNs7UmnrvflmQefY+m9i9j9i30D1hkY7Fu+E4BQTpCIJwR9szYl2PazPXxiYyEdzV3ommmrKL5+boaYlmIi6/oIBALB+UbxNXNYueFGGp96h8Dxbpw5WVyycgGlq6/IGGcYBq+feI3l05djtzh4p30P+/zmOXxf517ead/DFQVXnpGYVV2lK9YFQFSNYPS31/Uh2adxU5bXMemxCQSCs4dxCWoixVMgEAgmhtMVFkaz3bFdLRzb3ZJuPlF83dwxiRYCwWQQbO3hmQd/w5/U3I3dbabTvPVEI7u27k03RRlqu11b9wLQ9NoRLr19cCGro8lP/aZXB10nIbHo7TL2faQRzaIye3+/RkoGtH3QTjKuMuPy6fzJ5rt559n3ueiW+RnD4lqc935xgOLr5qS/gwKBQCAYSPE1cyi+Zs6Q65PRJD+v/AWvz3+Bpxc/xYaP/C1PvvcTZElGN3RkSebJ93/KIwUT51IbDotsYZpzGh3RDtxWFzbFNuRYpy8Lm9NGLBjD7skcp2s6sWBcpIIKBOcpE1ocZ9euXSxZsmQidykQCATnPeMRFkba7tiuFjqa/WkHTzyUYOvfPEdHU+ewjSgEgslm19a9XHr7gvTfLsDSey/n8U9tGfbv3lvoSf/tth/wDznu0tsXkF+SRzKaJL84d9DOtrrxAMFkEN9HfRnLf9f6WwyLARbTkeDw2Fn+qcUZY97teJd/f+bfWfbcTfzhp7sove8Klt1/ZpwTAoFAcL7xxo+2Ez0cp/Tw9XwQ3M1z+b9Ku9MA/n979x4W1Xnujf+75ghzYgBBUFQcNImHRAQTc9KcIObwpruNYNqm+81OW6E7TU/uFLVJ293dNAbjTg+7+9eCbZq8PSWCtk3SpgmYtok2aZURk6hJlPEEgiDDMMwMc16/PyazZJgZmAF0AL+f6/IS1lrPmjUPa+aGe+7neYJi8KJXqelVemiV2pjVaT2ubqQrNNCpQvNqKlRy6GZEx5lBmzu0wEHfIPQzdVBr4yfmiGjqmdA0eWlp6USejojokhAvsfCP5w6Mu92RV49GJCfUOhWW3bsER149BnvXAIhSpW3PqaiqrvC93N7aGatJUgx5eixesxDLPr4Ys6/KQ0a+Pupf5qwMzJs3J2JbMNOHxo4X8Ku25/Dlv3wR3oA35vl/dfg5zHm3CAAgBkVkfPRcLG+dws4Nf8LP1/0WOzf8CZa3To37uRARTWfBQBDeAR8AIKAMIP/aHBw+dwgCImf/FyDgF4eekb73DxlqeaHIBBnkQuS8aU6fE/3efnS5OnHWeTZuW1EU4bZ7pK/D8695HF70ne7HOYsVfaf74XHEjjNENPlNaEJNFOOPLSciotjGmlhIpJ3l76fw9nPmiGNyFoQm3m0/2DW+CycaI4/DC6/TC8NMXdQ+lVaFc5b4lWfJPk57ayd62noTbnOg+wDcfjcA4MoZy2IO8wnP69O66m18WPwudCYNilbNg+WtU2h68g2k6dUorlgCuUKGpiffYFKNiGgEMrkMZTWrcNt/3IhVVdfglqtuQZu9LaoyTISIY7ajeLfnHZw5dBbPfqYBnYe6L/r1Dl2VWavUxD1OEAQYCwxI06uRZkiDXCmHx+GFvWsAMrnw0e9sIuxdA0yqEU1RE5pQE2KtIUxERHGNNbGQaDvT9ZzbiSYf+9n41ZFpehXcA55xP0Z7ayc6DnYiZ0E2AODlbzUnlFi7de5tePrmH+KavGvw6UX3R+23e/qleX38ah/aSg7j3U+8DZlchgMN76GgOB93/eet8HsCOPvBOSjTldj3q9ZxPx8ioulMEAQsWF2IK29fhMajO85Xp4nAsjeuRe7pWdKxv37/V9j3y1YEvAH8MwXvr7maXORp8qFT6qFTRf6OFRSDCIpB6Xu5Ug79TB30H0074OobhEqjRMYsA/zeAPzeADwOD5y9rov6HIhoYkzoHGpERJScsSYWEm1XvnF11P7wvFMFy/KSuVSii8brHP8n9asfWiklk3N02Vi8ZiGat76JT9V9fNS2RcYiPHbtt6O2m8+24Hv/+C58QZ+0LSgGcdgamtfH1t6Pq+9fBs+AF4de/gAA4Pf6MdDtABAaDnqg4T3Y2vthLMjA8sqlI07STUR0qRm6sicAzP1gAQra5qOgbT6OXXkYH6w4iK73etB1pAcA0HW4G2fe7cKsKy/u7zQ6lU6aP22oPrcVA14HNIp0OPxO5GnyoBlSxRbwBaDWp8Pr8sLvDg1ZlavkCAZCSTiPwwtX3yACvgDkSjk0mekRU3sQ0eQyoRVqy5cvn8jTERFd8saaWBip3cFdh7DygeWsXKOUGWlSZvfAxAx7GX5/zyjKgr3LMeb52QLBAH7+3s8ikmlh4Xl9jAUZOG3uhKAQcM3/XY4Zpiyk6dXInGOMGA4675oCeJ1eDgclIhomXAEclt2ZK33dmx+ar+zyA1dBFELDQQWZgH2/eefiXmQcvqAP5wbPweVz4kPbhzjZfxI9gz3w+N1S7JAr5fC5fFKyTKaQQa6QRQ0HBQDfoBf9Z+wcDko0iY0rofb6669L/wBg9+7d2Lx5M9asWYNt27ZNyAUSEU1nY00sjLVdU+0bmL0sb8RVFIkuNLVODSB24tfr9EI1zlXQ3n7OHDW8M/yY9rOOMZ3TF/ShQDc75r7wvD7Zt+vR3tqJ/6mpw+m+U0gzqDFoc6Okcqk0HPTOb98C6wkb7F2h69j/24MAuJgBEVG4Om3okMkDN/8dB2/4B9qWHsG52V3I6sxF1tkcCGIo6eRRuPGe9R20v3MmVZd9ngj0us/huP04Bv0u6NU6dDrP4LD1CI73WyCKIjSZ6bDbB3DY8j7cQTfkShkgCNBkpkvDQXW5OoiiCEEmg9seWiUU4GIGRJPRuBJqO3bsgNlshslkAgCsWLECZrMZP/3pT7F8+XIm1YiIRjHWxMJY2h1+9SjUehVWP3TteC6ZaNzUOhVUWhXccf4YKFiWP67zH9x1WBraHOZxhIZBx5p3MBFpijT0e/pHPOYn/h/BcschBFxBnPjdGfTb7Fi5YRnmXzcXtvZ+zCnJh/3MAPraQ+fR52ox0OWIqF5b8emrkKZXs3qNiC45w6vTAAAC0H6ZBe9f3QoAWPr3FRG7lV4VZnTl4elXvg9PYPzzb46HUq7ElTOWYY5+DtIVWmgUGgRFEVqlBnmafFj6LegRz2JAZ0fXQBfOdHbC4/NAkxca1hnwBaDUKOF1eRFej0GTrUHAF4ioXtNkpUMmF7iYAdEkMK451EpLS7F+/XoAoeq0trY2NDU1obCwEPPnz4fFYpmQiyQimq7GmlhItp1l70l4nd6IZJrH4eW8HJQyRTfOhb0rci7A8PcFxeNLqK18YDkWr1kYsa39YBdUWtWYzz18Xp9YBv0uHMl/B/jYkOFHfcCL+KM0HPTKjy3Cp+s/jsOvHsXp/Weg1qsjFjP403++DplChsy5GTDveBem6+Zy7jUimvYSeY/N6syF3p4RsU2AABEi5hxcgN7DfRd9LrXh9Co9MtRG2Dz9cHgd8Af9UMvTIBNkOGk/jkG/G4IgIHdOLpxBO87IvdD6NbgMl0vDQTNmGaBQKeDud8Pn8UMQBKl6zZCnh/WUDSqNEoJMkKrXOO8aUWqMq0ItOztb+rqpqQlFRUUoLCyUtnHVTyKi0Y01sZBou562Xrgd3ohhnh6HFx0HxzaXFNFEKL53CY7/PbIC6/CrR7HqoZXS9x6Hd8TVOb1Ob8wqzRmmLFj2now4z8Fdh3DTF1dGHZuomJUTCdhQ+gjcfjeWVy5Fe2sn/vtLP8ae197CuWNW9J7oQ0nlUql6zdU3iPaDnTi1vwODNjds7f0R1WtX37+M1WuTUJezE//b+j/YebQRO4824s8nXpmwdokc87W/fBl7O/bA4XXA4XVIxxJNJb8+8svzK3vGcbn5Sojh0q0hBAhId2qi5lJz+pwTeo2JEEURIkTMM8zD5VlXYJ5hHkSI8Af9yEnPRboiDekKTUT1Wr52FoJiEIJexPHuE+g61Q2f24dgIIiANwBNZnpE9VrQH4Tb7kEwGMRgv5uVa0QpNK4KNav1/HCKxsZGlJeXR+y32WzjOT0R0SWh+N4l+OO3d+PaB0qkbbESC01b38DKB5Yjpyg74Xb2rgG0Nh5C0Y3zIhIM7Qc7sWhYBQ/RxWTI06OsZhXefs6M3AXZsJ91IE2vjqgs8zg86DlmhWfIvIAehxcHdr4Hr9MLe5cD7oFT0vnCSeOC4ny0t3bi7efMAICBLgdW/fvKC1qdFssC40LcMOtGfP61BzFLNwvBO+XI+mceTvzuDDIKDNJwUGPDezht7sSMouzQPDrWQSjTldDlaCOq19596X1cdpsJwUAQBxreY5XaJNDl7MTX/voVbC9/Rlrx79lDz2Dn0UasXVgxrnaJnrutvw21+7ZI368pvANfLP7ShD9XogvF7Xfjg74PYibLwrI6c5HVnRt3vwBBWvFzRlEW/vaHvXgm7af4xOJP4F+KPg61Iu1CXHr0dQgC5unnSYUlGeqM0HxoggC9Wo/3e+XodfdGVK8JgoAPre9jMODBubReKJ1KBN0BKNUqGPL0UOtUcPWFqtcEjQqCTIAYFCGTyeAbdEOblY6MWQYAgMMfhOD2w2l1sUqN6CIYV0ItMzMTX/jCF2CxWGC1WlFbWwsgNPxz69atqKysnJCLJCKazsaaWEik3c4Nr8Dr9MLy9+hqFs6lRqmWU5QtJYhjMeTp8eBv1kVsU+tUUhJ5pHu4oDh/3ENHw8LVaUMnyh5JvjYfWmUoAfJmxxvo8/Shz9MH5CHmcNDllUvR9OQbAICr/uUKtL15Cj3HenHdg6vx+vf34ur7l8HvDWDfrw/C7/ZDpVUi4AsAAIeDptjOo41YU3iHlPACgIqF6/DpP903YkItkXaJnntN4R0oMi4AABTnFCNPOzH3PdHFkqZIw7N3/HLEirK9/9UCq6wfGOFtOLzi55ySfFie78B16nK8tmo3sj+WjdvmlcdvOMGGj9IKfy+DDFqVFjqVDka1ETaPDSJEvNfzDnrdVniDXmTpMjEgt8Erd0Or1CBHF4qRmsx0aRRCujENLusggoEglBollBolAMA36MOgzQ33gAdBfxBZc40AQh9EcUgo0YUxroTa2rVrUVJSArPZjIaGBhgMBrS2tqK/vx9VVVUTdY1ERNPeWBILibSL1YaIEjeW6rROZycev/4JLJqxGHs79kCn1MHhi15d9KslG/D9lqdx1eyrMGd9Dj74w4c4c7gLWXMycfum1RHVa9oZWvjdfgCAUq2AIU8vDQctKM7H0v9zOXqOWtH05Bso37SaSbWLZE/Hm3hgyYMR28IJsNbuAyjOXT7mdomeO0+bjzsK7xz/kyFKoQx1BjLUGTH3nXm3C9YPRl4UBgDEoIiuw92wnuwDAKi8KmTM0uPmubdO6LWOVbzqNYjAoH8QHo8X2WkzYPcOSENBvQEvbB4btGot5NkCTnaeQpY8C1qNVloZ1OfyAcZ0aZin3+1HemY6AEiLGag0SsjVKoj+IOxdA1LlGxGNz7jmUHv99ddx/PhxZGZmIiMjA3a7Hc8//zx++tOf4vjx41i7du1EXScRERHRRZfIvD7DCRDwmw9+DaVMiZvn3ILn7vgV8odVDS0wLkSeJg9/Ob0bPzR/H7/Ez/GXu1/G4Yf+iWXfWgjtstDwpPDca0dePYrFdy6ENlsDp3UQJZVLpeGgd37rFhz96wl0Hz0HTWY6DjS8N2HPn+JzeB1w+pzI00RPgq5VatHW3zbmdsme2+F1oLX7ANpsx8b4bIgmr32/PghBltj7sCAToJ+px4KbCjH7+jxU31oNuSCX9ts67Hjj/TfQPnD6Ql3uyNcXo3qtMKMQhRkmGFR6uHwuaSioXqWHw+eA1d2Lk/0ncCZwBjaDFYF8HzT5asg0AjSZ6fC6fOg/Y4cgFwBRhCATkDFLDwDSYgbaGVp47B74vQEE/UFpMQMiGp9xVajt2LEDCxYsQEVFqOy8tLQUJpMJdXV1sFgs2LZtGx555JEJuVAiIiKiiymReX1iESHiA+v78PjdoXl7BECr1GGBcSFun7cGr518FQCwv3u/1MbpDw11aus/hg1//SoA4PNXVmHJoiUo37gKf39+H04d6oBhtha3rw9Vr4WHg559/xycvS4AgC5HC1t7ZCWHx+mFWqvi8NAJdtbVFXefXqnHgNc+5nbJnLu1+wDyNHlYllOMs64ufHPvo/i3JQ9Kw0CJprIz73ah60hPwseLQRG9Fiuu/2wJ8pfMjErEvf7jPej6oBt/uqIJ8++djarS6pQvpBeeY21h5mURQ0FFUYTL58IZRwcG/W4Y1UZolOnodJ6B1W1FmkKNfO0spM9Mg8/mR1+PDdaAFXMvL0C6IfShTMAXgFqfDo/DAwAI+oNQaBTStAFhwUAQvkE/h4YSJWlcCbXS0lKsX78eQGjeNIvFgqamJhQWFmL+/PmwWCwTcpFEREREF1si8/rEo1VqpUmwlTIlalc/BYWggCAIWFN4B/yiHzJBBrkgxwsf/Daq/YNLPoefvVsPALhyxlVwfcKFtv5jKMpYgGsXLUO3qxvGggycNncif+lMXHaLCSf3twMiYCyIHDb1h02vwuv0wtk7iNnL8nD1/ctw2tzJ4aEX2FhXGEyk3dBjvlj8sDRvmk61AHcU3onaf25B/e0/H9PjEyXK3jWA1l2HYMgLVUOptKqIeVxHc/jVo7B3DUQsrhS2c8OfsHztEhz8ffILwkAA9v3mHfzLltsjNnce7kbPYSvkUCDndD4CMr+UTGvtPoD6d+pQdVV13KHaF8pICxnM0s2CQqbA6YFT8AX90Cg0sHn6oVVqoFPqcG6wB/6gH1lZWQhk+GCzW5Eh0yPdnwaZTA65MrSYQZohDSqtCl6XF4IIyJTnq/YC/iDOHumBe8AN3QwtdLla+Fw+Dg0lSsC4EmrZ2efn7WlqaoLJZEJhYaG0bazZ/ubmZpjNoZW52traUFRUhJqaGmn/1q1b0dbWhtLSUmRlZUW1D1fMEREREY3HSPP6JEMpU0pfC4IApRD6/tNX3I+Ws/txzHZU2r/AuDBiAYR3z51fyGBoBdu373kC+3/wLgBgTkk+nFYXOg524fZNq6Xj+zsH0HcqVLGm0ihx93dugyAIuPJji/Cn/3wd+359ELOvyoNayz+YkqVVauPuG/ANjKtdMucevghBkbEIXa6uEedwIxove9cAdm54BZ+u/7iUcHn7OTNadx2SVlyO16511yEAQNueU1i0JnYl5bk2K5q2vjm2ixOB7g964PP4oVSf/3PXMFOHRXctwJHXjqGr9CS+sXgTul3dsHv78dyhZ9Fl7cQv3nsGXyr5MgyqDORq4q8qOtHiLmQgyJCnzYMn4MFJ+8mI1UHx0XQEZxwd6PfYISIoVbCdGzwHuSDgMuMiOM+GKpiVGiXEQBA+tx+GvPPvMV6XF4N2N2QyAdpsDTTGdMCYjv4zdrj6BiFTyqBQyVNeyUc0GY0roWa1WqWvGxsbUV4euXqKzWZL+pxmsxk2my0igVZUVIS2tjbU1dUBCCXZ6uvrY7YvKSlhQo2IiIimBL8YWmhg+HDQVQWroVVq8cfjL+Ok/UTMtt/p/wZuvPcWzDiYhX2/PghjQQZu37Qa/UW9eKj5CVRdVY3ZjrnIW5yDrsM9yFuUI/1BJAgC5pTk461fmPHs/TuQXZiJTzx1B+QfVS1weOjodMpQVU6sijKnzxk3KZZIu0TP/eyhZ7Bq9uqI4Z3htl0jDBslGq/WXYewaM2CiOql5WuX4tn7d4yYUDPk6aUVmnuOWeMet2jNAswoyoZv0IcZpizocuInmWNRaZQRyTQA0GZrsLr6Wqy4bxnkmkqoVWp87Pd3AwCUbhVu3fkvODunA9889U04jQN48eN/TOoxLxRRFCFCxDzDvIghoVlpWUjzpyG0rkEQ5wZ7pQo2hUwOAQLOBNqhz8xA0KWAyzoIuVIOQ54eAbUflv525GnyIIMcAgBFmgKqIR+uKDVKOM+5YDvdDwhAml4Nfe75VYe5eijROBNqmZmZ+MIXvgCLxQKr1Yra2loAoeGfW7duRWVlZdLnrKurQ3Nzc0RSrKysDPX19VJCDQi9sQy3detWJtOIiIhoyog3HFQpU+KO+XdiTeEd+I+/fS2igm2mJk+aYyunNAtrH7gLAKRKi//9549x1tWFZ9/7BR4u+RKu++Zy/G3zP+HzBKRhRKIo4uS+DoSnhwsGghHJtKYn30B2YSZmXTkTrj73BR0e6vf4oVAn/itpssdfKDqVDlqlNm41WnFO7OqwRNoleu5dR3ciT5sfkVBzfNQm1oIGRBOlbc8prHwg8h4PJ1PaWztRUJwfq1nCDHn6pIaPJkNjTJe+3lD6CJ5u2Yb5hy+H0qtCQdt8BJR+XPGACUBqh4KGjTQkNE2Rhqy0LJy0n4Q/6Jcq2ILBIPTqUHJdZ9BCmx1KSPqCPgQCfhzvP45uVzfkghx5mpnQF+gghwzKtPPvrT6XD2IwtMgBRABDKtTCq4cG/QEoNSoE/QEOEaVL0rh+G1m7di1KSkpgNpvR2NgIg8GA1tZW2Gw2VFVVjemcw6vcEj3GbDbDaDTCZDKN6XGJiIiIUiHecFAgdgWbL+DF/Ys+g4PdB7Fi5grp2M+/9mDEeS32Nml46A8++RM0PfkG/vSfr2NOST5Omztx5t2zmFMyC4P9buReNkNqF149VKYQcGr/GQBA3uJcHGh4T0qoDdrd8Dq8MOTpIyb9Tray7chrR9G66zDu+W5ZQhUojh4nXvpmM4rvXYxFt1+YP7aTcePsVehydkZsC38/0h/fibRL5JgHljyIOwrvjDimtacVWqWWwz3pgvE4vPA6vTDM1EXtU2lVOGexjjuhFn6cnmO9UOtVyCnKHr3BGNxUcDNebPsDvGoPvGoPFF4lTi87hi/P/Xccsx3Fzj/8AfP3LsFb9nfxbn4bllZejuU3X3lBrmUk8YaEArEr2Nz+QWSojHAHBpGuOJ9APN5vgdXdh07nGcgFOT6wvg+r2wqVXIVcdx76z9ih1Cjhc/ngdfmgzdYg6A8NEx2abHP1DUKZroDfIyDgDVWoqdQyuPoGpYRawBcABAFyhUxqx6o2mm5kox8ysvnz50MQBKxfvx5XX301tmzZgr6+PqxduxZr165N+nwVFRURlWhAaDXRcPVb+Jjh6urqxpzEIyIiIpqMwhVs/33T93HH/Dvx3zd9H0/f8kPcd/mn8MSqJ3FlzlXSsRtKY6+svqH0EZium4vyTavR2dOJt35pRl+fDWWbbsRd374VRZvy8duiZ9DafQAAYGvvR8HyPJx9/xyAUNVJ4crZEauHHt97Cs//+4v4xadeQNvekwDOV7al6dW4+v5lSNOr0fTkG7C8dSrmdfk9frTuOgx75wBeeqwJjp6RJ+N39Djx0mNNsHcOoHXXYfg9/sQ78gJZu7ACf+/YG7HtzydewUPFD0vfO7wOfHPvo2izHUuqXSLHFGUUYW/HnojH2vVhIx4u/vL4nhjRCOxn488RmKZXwT3gGfdjtLd2ouNgJ3IWhBJpL3+rGT1tveM+73DhDy0UN4go+k4+Tt3zAfS5elQ1fw5bn92GgpcXQN9nhCwgg7N9EP/8/kEceuN9AKHqtYeavyC9d6ZKuIJtnmEeMtQZmGeYh8syL0eOJgdz9HMhE87/yZ+vnQW5IIMoikiTp0EuyKFVajA3dy4MeXoEAyJOd7Sjf7Af6lwVtNka6GfqkDZLhTPBDrh8ofnYAr4AZEo5xGCozFmukkOpUUasHursdcF6og+9J/oQ8AWkqjaZXIAmKx0yuQB71wA8Du/F7TCiCTTuevl169ahubkZZWVlmD9/PkRRRE1NDRoaGvDqq6+O+byNjY3Yt28fzGYzGhoaUFZWFvfYjRs3RiTc4vF4PPB4zr/B2+2xlzMnIqJLD2METVYjVbANdVPBzdh1dCdO2I9L2xYYF+KmgpsBAPOvnYO/2V5Bv8cGjUKDsqtuwDHbUTx36Fm0O0/jF+89gy8ufxi6WRqcNnfinsfL0HvcBo/Li1P7OiJWDz13vA8A4HP7ockMVT+EK9tWf3El/vSdvyCr0IisecaIyrahFGoF7vluWShJ1uXAS4814Z7Hy2NWqknJtC4HDHk63PPdskkx7DNPm4+N12zGs4eewULjZehydUGvMkRUjTl8AzhmO4oB70BS7RI5pjh3OVq7D+DZQ88AALqcXXio+GFWp10gjBOJ8TrHnyBZ/dBKafXQHF02Fq9ZiOatb+JTdR8f97mHGj7s/s4r7oRf9GNvxx7s+/17ECFC+GjyfwECRIh4p+F9qK+SR1SvHZplweKKy1JSvQaMXME2lF6lh1GdiR7XOQSCASgUSqjladCr9IAKENIBuR7wwg2HbADpwTQEgwH0DPagfaAdAgTM0uZDlIsI+gIwzjbA7w1AppDB3e+Wpg4AAL83lFwLBoKQKWRwdTmg0iiRMcsAj8MDT0CEb9AH+9kB5OguTAUi0YU2rt9Etm3bhvvuuw87duyI2rd9+3Zs27YNjzwS+9PS0VRUVKCiogKNjY3YuHEjGhoaYg7ntFgssNlsMBqNo55zy5Yt+M53vjOm6yEioumNMYKmOr/oh0KmiFrgwC/6oRSU6HR2ot9jAwBckbUIVU2fi2h/3G7BI3/bgLzCApT+ZRXefvaANDy0vbUzYvXQ3MuyMWhzo/e4FdmFRgChyrar718G6ykbbO39sLX3Y9aVM9H94bmIx2l5/h2kGdTInp+FvEU5uOfx8hGTalHJtDhJt1QpMi6ImMNsuDxtPn57d/TvyqO1S/SY4tzlTKBdJIwTISOtCuwemJhqo3AyLWxGURbsXY4JmZ9tuFgfWtxUcDOO2E9LyTRpPwT0dfRj67PbUPqXVVLCbeC0C//8/kFkqDMm9QIuoihCKVdiyYwlEQschOdkc/sHpWPTFWk43m+B0+eCJ+CGRpmOd8+9g/etR6AL6HCFb0lo6KdGCXe/G16XL+Lnptaq4FeEKgAFQUDAF4BaH/oAxuf2w+/xI+APQhhWbRwMBOHqG4RCpYAiTQGFSg6iyWpcCbWMjIy4wzrXr1+Pp556ajynBxBKrO3btw+lpaU4fvx4VOKstrY2oXnXAGDz5s3YsGGD9L3dbsecOXPGfY1ERDT1MUbQVDfSAgcAMEs3Cz+7/Rc41PseDCoDbp5zC55u2RZ1nrOFHbil5jq8u/MDafXQBf8+G78WfoHCw/OxqmA1rihbAPdVA/jDO7/FYkchijXLYSzIwGlzJ5TpSsgUMgT9Qbj7PRGVbQF/EOaG9xD0B5E5NwPr/uce6HK0EUm1xq/9ERpjOga6HTDM1GPQ7sagzZ2yZJon4IFarr5gx9PUwTgRotaF7u9YlWhepzdipcixePs5M4punBcxb1r4Me1nHeM6d6L8oh/eTDfU59Ijk2oCoMlTY2Hr0pjVa/t+2zqpE2ojLXAAAAZ1BtSKNAz6B6FRaKCQKdHpPINB/yDS5GnwB/3QKbXIzZgJg0oPV9+gtHpoMNOPPsEKtVsFgyoD2mwNXD4XulxdUPoUkCvl8Ll8gDEdYiA0VNQ/pMo5zO8JYNDmBgCkZaRBP+w93zfog8/jh2fAy7nYKOXGNYdavFLSsMzMzPGcXlJeXg6bzYb6+vqofTt27EBJSUlC51Gr1TAYDBH/iIiIAMYImh6UMqX0+5kgCBGVFwCQq8nFLXNuRenMFbip4GYsMEZO7K+QKTFbV4DLbijC2qfvwud2fAprn74LZ+d14J9d/8SOD1/A4d7D54eKOkJDRZ9572dQ3ypDe2snLHtP4erPLMPMK2bAesqGksql0vn7O+wI+oMAgKx5Rml7OKkmU8jgGfCi73Q/TDfOQ3/XAAZtbqQb01KSTHv1xJ/x5dcfRo+rJ6Hje1w9+PLrD+PVE3++wFdGqcA4EaLWqaDSquCOM/dVwbLxVZAd3HUYPcesEds8jtBQ21gLIVwISpkSd322PJQwC//JKwAQgdX/ej109oyY1WsDZ84n/F7/89/wg8/+FPVrf42Gr7wcdz7Ji2204aFquRpGtREquQp6lR7qjxJpA54ByAU51PJ0ZKVlQq1TIXNOBmaYspA5JwM+lQ8O3wB63b3wBX3w+N3SUNEzjjNwpTlg67ejr8MGuVoOZZoCaQY1dDMi39cD3vPzsMWqTutps6L9QCcGuh2ci41SblwVaseOHRvX/lgyMzOxefNm1NTUSNuysrIAAG1tbRHHNjc3w2azcWVPIiIioiTFWkEUAB6/8YmoYzscHdLXPzn4vxH7jtstOG63AADK190Fd0sWWn77DowFGbh902roizUIiAHIBTkMeTr8y5O3o/eEDfphfxjrcrSAeP77D3eHzilXyZGekZaSyrRdR3ei03kGj+7ZhO/d+CRyNDlxj+9x9eDRPZvQ5erCrqM7cfOcW1ipRtNW0Y1zYe+KXJwg/P14h2SufGA5Fq+JTPa3H+yCSqua8OGeI1l4w3zIN8lhfv5d2Dr6YZydgdJPXomClflxq9eMBRnodnXj2N+P4+hPTiMNWogQYT1pQ9OTb6B80+pJXcE23PAVRGd5ZkOEiOy0GVHH+oI+AIBcUODUwMmIoaInBo7D4/dAJahwhWcJAp4g5Eo5suZlQqVVwhf0SR8AqfUqyFUy+L0BKNMjPxQK+AIY7HdDmaZAxiwDNMZ0wJiO/jP2iBVGJ4suZyd2Hm1EnjZ032qV2qiVmcfaLplz//nEK+hyduLflnx2PE+HYhhXQu2+++7DmjVrsGnTJpSWlsJgMMBut6O5uRlbtmzB9u3bkzqfzWYDgKgEmcUS+oWqtLQ05nYiIiIiSs5oQ0SHemRFDf7v4n/DcbsFg75BfN/83zHPOXflLPzL/XdFbHvglc/A6XNiyYyl+M7130XeolzkLcoFEFolr/6dOlRdVY1lOcUARBhm6WE/c/4P9SvKF+CD5uQ/pB0vtVyNx294QkqSjZRUG5pMy9Pk4fEbnmAyjaa14nuX4I/f3o1rHzg/Uujwq0ex6qGV0vcehxdNW9/AygeWRwzfDPM6vTGHjc4wZcGy9yRMN8yTznNw1yHc9MWVUcdeaKbr5sZMgN312XK8vnWvVLUGhP4vve9KfP61B7Hqd3dAD+P5hJsIiBBhfv5d6XxD3/8m6zyIow0RHarQUAhv0ItAMICAGJCGimoUGnS7epCuSMMMfQ5yM7OhVqRJ7bwBL04NnAx96KLKQHZ6NlQaFVSa0P7wsNE8TR7UQhpU6UootUoo08+nMpQaJVzWweGXlFJdzk587a9fwfbyZ6BThT5AevbQM9h5tBFrF1aMq12ix+w82ggA2NPxJtYU3nHBnuulbFwJteXLl+PrX/861q9fj+PHz68oZTQaUV9fj+Li4qTOZzQaUVVVFTWEs66uDiUlJaiqqorYPrxijYiIiIgSl+gKooIgYKZ2JmZqZ0IURbxkeRHHbEel/Tnpubgm7xosmbE0op3NY0OfJ7QiaCB4fhhPt6sbdm8/nm75b9g8fdi670k8du23oJ2lgfOsCwDgTh+EO92FQ7vfR2aecaKeclJyNDn43o1PjphUG55MG62SjWg6MOTpUVazCm8/Z0bugmzYzzqQpldHVJZ5HB70HLPCM2ShAo/DiwM734PX6YW9ywH3wCnpfMX3LgEQqnBrb+3E28+ZAQADXQ6s+veVF7U6bTTxqtfmXzcXG04/gsPPnYo5JNTW0Y9uVzfOHO/Eb3a/gPnmybFK6EgSXUFUEITQBwkfjdK0ukNDRR1eB1QyFXI1echJz4Fq2IcN3kBoOG9ADESc2xf0IRgM4HDvIXQ4zsCRMYDLMy9HWrYaKqUSGmM6vAEvRIhwOQYhU4xrNqsJt/NoI9YU3iElvACgYuE6fPpP942YUEukXSLH5Gnz8cXiLwEAjvUdBV0Y415vvKysDMeOHYPZbEZLSwtMJhNuu+22MZ+vtrYW9fX1aGlpgdFohMViQUlJCWpra6OOLSoq4nBPIiIiooso3lDRz175+ajqNo/fg+tn3YAT/ccxP+P872yff+3BiOMcPgc2vVmDvAUFKD29CnKVHMZrtTj9gQ2wCjgpt8DR45SGffoCPjy1vxYGlQELjAtxx/zIYS69g72Qy+TQK/WQy+TjqgSJlVS7f9G/4oUPnsd9l38Svz7yywlLprV2H8Av3vs5Hlz6uajrHGkfUSrkFGXHrDwLM+Tp8eBv1kVsU+tUUlXb6oeujdu2oDh/UiXQYolXvXZTwc04mLkdst7YQ0I//9qDuHXHx1DkXDrlVglN1PChouHVRHM1uVEJOZkgh0ahgSfghUp2fsjm8X4LBrwOnHachlyQ4/TAaQTEIFRyFXJdeeg/Y0cfrLDbB+AfDOCyoshhwh6/G32ePsgFObRKHTRKTcR+X8AHmUwGuRDKAA6thBt+7FjsaX8TdxfdA5fPJZ0vnABr7T4Q9318T8ebeGBJZIwMt/tH59vI187CnvY38cDS2Me8duJV3Dh71YQ8BxrduBJq27dvh9lsxk9+8hOUlJQkvDjAaIZXoo10XKLHEhEREdH4JTNUdKZ2JjZd8w0AoT+wwjaUPhJzhdG8k3OQbkxDekYarHtsUBpU2H/rm5hzdD5eeqxJWpjA7rXj7c63AAD9nv6ohNqPDvwAB7pD1S2P3/CEtIDCc4eeRYejA/u79sGgzsDt826PqKoLiAHsPtkMtVyNGekzpH3hpFrN3/4DXa4u/KDlaQQQwA/N34df9CMnLQePXfttuAODON5vQXZaNgzqjIjzvnXm7/AFfNCr9FiRd3XE9b7c9hKO2j7EgW4zbB4bfnXk/0Gn0sGgysBZ11n87fRf0HJ2P3rdvRH7cjW5Sf3siOjC84t+9F53BrNfXiANCRUhQhAFlN53JfSKr+CEsxsAIlYJhQCYn38Xs6+ciQ5/B+ZlzINckE/JZHoyQ0U1Sk3M5E++dhZ8gRMQRRFpijT4gn5olRrkG2dB5VWHVhjtd0OUiUjPVSNdnxbR3hv0weELLRKhlKugQeRjnBo4CREiZIIcs7WzpAUU5IIcaV41vEEfVDIVZqTPgFx2fnEEX9CHQZ8LMkEGlVwNlTxy3rZB/yAc3gE4/U6IooiewR7kCTMhk8khQIBWqcWH1g9wVc4yyITzVXX+oB/nBs/B6XMiW50VcU5f0AeNQoP9XftQnFsMpz90THjuuX5PP1w+J9Lk6bD0t2Fx9mLIBRlkMnnM2EwTZ1wJtaamJlaIEREREV1iEh0qOtTQP6RuKrgZL7b9IWLYaEZPFq5wLcHHtt0OXY4WLWf3459d/8B8+xxknsqF/bRDSqoNqM7PsaZX6aMea8B7fv9je78hfd3Wfwxt75yfj604pxhLhrTz+D34ceuPAADLcorx3Rnfk/blaHLQ6+kFAAQQGr4artbrcffgw74P8KMDPwAAPLTs4agk39Z9TwIArshaFJVQq3/3p9LX4aq/DX/9atTzuiJrEd63HpH2vfjxP0YdQ0SppZQp8c3PPYrTS87A/MJ7sHX0I2O2His+uQzzr5uLrDNGHBf+AEEcllwSAVtHP17+r9043X4afXN7kLM4C8G/KXB539V4I3MfXJ/yYsH186dEMj3RoaLx6FV66FQGZKfNgFquwqDfDbU8LfSerwpVOwq5IrwBL4IfLXwzVEA8P83A0MQVAATFIMSPJr/rdHSg39MvLaDQ6TwDp88FASLmGuYhKz0Lcpw/t8fvQfdgKCGalZaNLHlk8uvtzrdwsv8kAEAtV6HTeQZWtxVapQYZKiPS5GnocHbAH/RHJOO8AS/etx4GAAwG3BHnPGk/AaVchV53L/o9/QCAnsEenLSfwALjQhy1fYgB7wDUchV8QR/cATdOO04DQNRq3jSxxpVQu/rqq/H1r3897v7Nmzdjy5Yt43kIIiIiIppmwokok64I2W/n48OZhyFXynDXf90qDessnbkCpTNXAAAcVzjx0mNNsHeFkmp3ffdWPLPmOQx47UhTpEed/8oZV8GoNkp/cMSjVkTO5eMZ8kdMmjxt+OGYp5+HkwMno7ZvKH0kYmhXeLW7MLkghwwyBBGEP+iPar84ewkO9x4CAFybf700jHZD6SOw9Fvw+2O7AADzDSa8bz0i7SOiyUkpU8J0/TyYrp8XtU+Tlw5P9mDMVUIN+Xqc+9CKdFEL4aQA5RGNNCxUdi4d7/7PUTz77jP436/8eNpXHoWHjRZmFEYMGx1a6WZUG+O2N6gM0Cq1CAYDUAzrKxEidEo9gmIABfq58ATc0gIKNk8/0hVq6JShD2tkkEW1DRueqAOA7LQZOOvsBACo5WkIimKosk47C4P+0MIJg/7BiPMAkQnHyD3ATE0egNDq00NjXnj7jLQc6YOkoTEmvJ8unHEl1CorK7FtW6hcv6SkBFlZkdnZ8GqfRERERDS12LsG0LrrEAx5oT8qVFpVxITj42l39p1zuPfYp/Hhbgt8Lj/mGIrw8W1rYJyZEeuU0OVocc/j5VJS7U/f+gsqf3g3ZmTMiHn8g0s/CyD0B9l//O1rEZVwpowiPLrym/AEPMhKi/zdNU2RjoeLvwxPwBNzLrQrZ1yFUwOnIv4QUggKLM5aApu3D7fNLYNCpsBcQ/QcSJ+7cj1kggyZ6syoff9+1Rexdd+TOO04hf96+9sAgMszr8BNBTejOGc59nf9E+2Odrxy4o8R+4ho6lHKlNGrhH70/+I7FuLQ3g/Q974dsoBMSqYBoWGhIkQs/9sN+F77VmRW6PCV674Sdf6pOEQ0lmSGjcYiE2ShhFeMxKNckCNPez7ZdNJ+UlpAwR/0Y45+Dgr0BRBFMSppliZPQ256LoIQkR7jg5ccTQ6ynKHYNOgfhD/olyrrRIgY9A9CLVdFnVcpU2KWrgAAos6rV+nhDXiQrkiTEqkquUqq0J6lmwVf0AtPwAOXzyVdZ6wKbppY40qojTbcM9myTiIiIiJKPXvXAHZueAWfrv841LrQkJS3nzOjddchaSW+sbZrb+3EOYsVN37+GmTPzcSBxkMQgyLe2XlkxEnKpaTaN5tRfO9iKNSj/xobbwEFY5oxZnVHuiIdtxeuiXmuHlcP9nX9EyJEKAQF7in6GF5qexF+0Y9H92zCE6tq8ZWSr8W9lnuKPhZ33yz9LOjVelyuvAK3zS3D7lPNkMvk8It+aFVaGNQZuFypi9qXyFBbIpp8RloldOndV8A36MOzn9mB4PAqJggQRAFZ7TMhqhwR+/7xshnvvnQEvp4ACgyX43c3vghdxdSeb3G8w0YTEW8BBYWggCCLfjylXAmlPPaHPwCQp8lDnz60unV2WjbmGeZJlXUGlQEuvwvZ6TOiYpBCpsAs7SwA0VXOohhKxGWojSgyLgAAuHyDUoJRKVMiXaGBO+DGjPScjyqshaQSkDQ240qolZSUYPfu3cjIiH1DrVu3LuZ2IiIiIpq8WncdwqI1C6SkGAAsX7sUz96/Y8SEWiLtjrx6FOUbVwMAFt2+EAtvmo8P/3ocb/5//0DxvUukyrZYdDlaVP7w7oSSaUByCyiMpMfVg0f3bMLZwbOYmT4T37vxSeRqc3H3/Huk7Y/u2TTmVT6VMiW+e8P34l7nSPuIaGqKt0ooACjTlciYbYD1pC1iWKgIERCAgTl9uCF3hbTd8tYptG4/HJpkHzLo+4wwvJSJrY5t6Cps53yLIxhvJVys8y3KWgStUosAAqGE2rDzFefErhzUqXTQKrUY8A1EbA+3XTV7NWbpZkGr1EIlV0nbBUHAbN1sAMDK/JUo0M9hMu0iiR70m4Ta2tq4yTQAqK6uHs/piYiIiCgF2vacikpshZNk7a2d42pn+fspvP2cWdqvUCuQsyA09LL9YNeo15ZoMi1MKVNG/NEx1mRal6sLeZo8PLGqFrnaULVHrjYXT6yqRZ4mD12uLjy6ZxN6XD1JnT+R6xzvcyCiqWfFp5ZJq38CQOhLAeWPrMLXHn0YqwpWS8fu/+3Bjw6JHB668OBSbCh9BOcsVvzyV7/FT3b/BK+f3A2P3w3LW6fQ8JWX8bOK3+D/fbEBj9V/E63dBy7205wUJroSThAE3Dh7Fbo+mkstfL7w9yMNxR3aLmx4uxtnr0KXK/KYs66uiGOYTLs4xpVQu+2228a1n4iIiIgmF4/DC6/TC8NMXdQ+lVaFcxbruNqZrp87YhXaZDI8mRarAi1Hk4Pv3fjkhCTViIjCTNfNRfmm1cielwm5UobseZm4fdNqFN1YCE1mesSqlv0d9qj2AgTo+jNwU8HNOPq343A1BCD7kR7Pv7gDJ9/uQNOTb8B60oaALwhXuxtz/ng5fvfiizhmO4puV/eEPx+/J3pBlok8frJZu7ACf+/YG7HtzydewUPFD0vfO7wOfHPvo2izHUuqXSLHDOX0OeH0Ocf8XCi+pD7ie+qpp2C1Rv4SNXTRgZ07d0IQBNx7770Tc3VEREREdFHZzw7E3ZemV8E94BlXu/Bwz6F6joV+vyxYNnlWJPMEPHhs7zdGTKaFhZNq4eTbY3u/gR/d+mOo5eqYxxMRJWKkYaFDGQsyYD1pi1geUoQIX5YHftGPziNnpe0ZRTocbDhyfkEEnK9oy/hrLjbM+CoATOgw0SOvHUXrrsO457tl0krOI3H0OKX5MhfdPvpiOJNRnjYfG6/ZjGcPPYOFxsvQ5eqCXmXAHYV3Ssc4fAM4ZjsqrdCZaLuEzu11oPHoDjh9TnS5ujDQ8abUdu3CiovQA5eGpBJqVVVVqKyshMViwaZNm1BZWRmxf+3atTh+/DieeuoplJeXo7i4eCKvlYiIiIhSzOv0Tni7g7sOYeUDyydV5Zparsa9C9di19GdePyGJ0adGy2cVHts7zdw78K1TKYR0UVT+smr0PTkGxGrhgqigDs/WwalTInr/m0FTr3Tjo7TZ7Ci9FM48MyHGLbeAQQISHeEkl0bSh8BAIhBEU2//ive9P8VuQuycdMVN+GqnGWwvHUKLc+/g/4OOzJmG1D6yatgL+pF/Tt1qLqqOmJIo9/jR+uuw7B3DuClx5pwz+PlIybVHD1OaUXn1l2HsfCm+UkP9Z8siowLpEUEYsnT5uO3d+9Iul0ix+hUOvzbktCK118s/lKCV0zJSurOzMjIQElJCV577bW4x8yfPx9f//rXsX37dphMJhgMhnFfJBERERFdHGqtKu4+90D8pNhY2zXVvoHZy/JGXOwgVdYU3oGb59yScHIsR5PDyjQiuujCw0NjrRoKAPmLc5G/OBdACQDAMrsrqqINAJx6By4zXo6bCm4GANjPOnC8sQMFWIiuOe3o+I8O6I5lRiTvek/0oenJN3DiziNozzuN5w49C53q/OqiCrUC93y3TEqSjZRUG5pMM+TpcM93yyZ1Ms0T8CT1fp/s8TT5JTWH2rZt2/CNb3wjoWPXr1+P+vr6MV0UEREREaWGWhf6ZT9WRZnX6YUqTuJsLO0Ov3oUar0Kqx+6djyXfEEl+8cP/1giolQwXTcXFT+8G59v/DQqfni3lEyLpfSTV0mVbMBHq4cCGFjVA4VcAb8Ymr9s6JyZ9uw+FBoK0fL8OzGHi2b9IzRkv63/GDb89auo+tPnsPtkM5pOvgYLjuHu794GQ55OSqo5eiLn9IpKpo1SyZZqr574M778+sMJz5nZ4+rBl19/GK+e+PMFvjK6mJJO9yZTcSaK4ugHEREREdGkodapoNKq4HbEriorWJY/Ie0se0/C6/RGJNM8Dq+0KigREV0YwyvaMmYbsOKTV6Hw2vvhF/3SSsJ5i3Jw69duQFdbN5YvvQImYxH2drTGHC6q7Y/ME3zq1GfxwfPtGDD04/Atv8Kzn3kO9zxeLiXNnvnOr5Dm0QBWAbqZWvidAQza3HGTaa3dB2IOKU0FT8CDXUd3otN5Bo/u2TTiHJtA5AI3u47uTKrymSa3pBJqbW1tSZ18+AIGRERERDT5Fd04F/auyEUGwt8XFMdOqCXTrqetF26HN2KYp8fhRcfBTphumDfu6yciopHFW/BAKSilr7VZGiy8eT4W3jxf2pYx2xBzAQRHxvmVRhcYF0Lbr4fN54SxNxtGoxGCIECXo8U9j5ej4SsvA6e1CEKEAMDe4QjN4WZMwz2Pl8OjG8TPzfXI1czEbF0BZuny8dyhZ9HuiB5SmgpquRqP3/CElCQbKak2fLXox294gsm0aSSpIZ/JJshsNltSxxMRERFR6hXfuwTH/34qYtvhV49i1UMrpe89Di9e/lYzetp6k2pn7xpAa+MhpOlUsOw9Kf37x/8zQ5+nu0DPiIiIJsLw4aIQQhVqjht68dCyh7HAGFqVMz07DaoZSsg0Au5aepfUXpejhVwh+6ipEPG/WqeCLkeLdkcHdp9qxm/f/zW27a/Fhr9+FW39xwCcH1L6+dcexI/MP4i6vi5nF473W2B1WxEQAxH7WrsP4KHmL6C1+8C4+yG8EE2eJk9Kqg0f/jk8mTZaJRtNPUlVqGVmZuL111/HrbfeOuqxra2tHPJJRERENAUZ8vQoq1mFt58zI3dBNuxnHUjTq7F4zULpGI/Dg55jVniGLDiQSLudG16B1+mFZVjiDcCknkuNiIhiL4Cw7L7FWHDd/RAEAWsK7wgNG705VOkWDAQhk0fW8cRbqGbgrAMA0O06K227dc5teP307pjHH7V9GLXt98d+hz8dfxkAsO2mp3FZ5uXodnXD7u3Hz97djnbHaTzdsg2fvPxTWJF3TUSV2xnHGajlauiUWqgVaVHnHj7sNJxUG1qpdv+if8ULHzyP+y7/JH595JdMpk1zgphE1uv48eNYt24dXn/9dej18Zc1t9vtuO2229DQ0IDCwsKJuM4Lwm63IyMjA/39/VyNlIhompio93bGCCKi6Wci39sZJ4jG5oUvvghbuz1qu7HAgPv+92Pw+N3odHbhrKsL8/TzsHV/LY7ZjkrHySBDEEFcNWMZHr/xiYhzbN33JPZ0vAkA2F7+DGZqZ+Jjv7877rW8+PE/Sl8/8Mpn0OfpQ256Ln625hfS9m5XN147+Sr+2PYSnH4n5urn4aulX5OGnR63HcdjezdjwDcAOeQIIACFEFrcISctB0+uforJtCkkmff2pIZ8zp8/H5WVlSgsLMTPf/5z2O2RLwK73Y6f/exnmD9/Pu67775JnUwjIiIiIiIioovH0eOEJ87iNR6HF44eJ9SKNBRmFGJl/rXI1swAEJqXLTyk1GQswvN3N2DTNZujzlGcsxy3z1uDa/JWwqjOAABsKH0k5uMN3+70hVYe1SgjF0T4/GsPYscHz8PpD+0/NXBSGnYKAIOBQQz4QvOFBhAaZhpeKbXH3cNk2jSW9CqfNTU1MBqNWL9+PaqqqmA0GpGVlQWr1SrNmfbTn/4U69evn+hrJSIiIiIiIqIpyNHjxEuPNWHQ5ka6MQ1qnQoDZx3Qz9TB4/Bi0ObGS481RazyqZQpUbv6KSgEReSQUpky5mPcXrgGtxeuidh2U8HNeLHtDxFVbvMMhbip4Gbp+4AYwKqC1XD6nMhJj0yAbSh9BE+3bIt6rHBCzhNwx33O8ZJ5ND0knVADgKqqKpSVlaG2tha7d+9GW1sbTCYTKisrsXHjRsyfP3/0kxARERERERHRtBdOptm7HDDk6SKSZsP3x0qqhQmCELESaSLC1WILjAtx+7w1eO3kq9L28LnkghxfKflazPY3FdyM3x/7HSz9bdK2BcaFUkJulnY2Pnn5p7DjgxcQRFA6RiEosDhryfDT0TQypoQaAJhMJtTV1U3ktRARERERERHRNDJaMg0Irf55z+PlcZNq45FsldtwftEPmSCLm5CTCTL85dTrCCIIhaDAPUUfw0ttL8Iv+vHonk14YlUth31OU0ktSjDdcCJRIqLph4sSEBFRPFyUgCiSvWsArbsOwZAXWnRQpVVFrMw8nnbtrZ041dKB95vb4HP5kGZQY+3Td42YJItIvuXrUfnDu6FQj7kOaML4gj4pISeKopSQ63H1SKt8zkyfie/d+CRytbnodnbj0T2bcHbwLFf5nGIu2KIERERERERERDT12bsGsHPDK1j5f0tQfO8SFN+7REqUjbdde2snzlmsuP5zK3DdgyXQz9RBoVbA3PDuiOcOV6oZ8vUovnfxpEimAaEqN0EQAHw07HRYMi1Pk4cnVtUiV5sLAMjV5uKJVbXI0+Shy9WFR/dsQo+rJ5VPgS4AJtSIiIiIiIiILjGtuw5h0ZoFUOtU0rbla5fiH88dGHe7I68eRfG9ofnDFt2+EOv+5/9geeVSHHn1GOxdAyOeX5ejReUP78ai20evlEuV4cm0WBVoOZocfO/GJ5lUm8aYUCMiIiIiIiK6xLTtOSUN2QwLJ8naWzvH1c7y91N4+zmztF+hViBnQVbomINdo17bZKlMi8UT8OCxvd8YMZkWNjyp9tjeb8AT8FzkK6YLhQk1IiIiIiIiokuIx+GF1+mFYaYuap9Kq8I5i3Vc7UzXz41Kuk0Xarka9y5ci3ztrITmRgsn1fK1s3DvwrVQy9UX6UrpQpu8aV8iIiIiIiIimnD2s/GHXabpVXAPxK6iSrRd+cbVUft7joWSbQXL8pK51ElpTeEduHnOLQknx3I0OfjRrT9mMm2aYYUaEREREREREUm8Tu+Etzu46xBWPrB82lSuJZscYzJt+mFCjYiIiIiIiOgSotaq4u5zD8RPio21XVPtG5i9LE9aqIBoOmBCjYiIiIiIiOgSotaFqqViVZR5nV6o4iTOxtLu8KtHodarsPqha8dzyUSTDhNqRERERERERJcQtU4FlVYFtyN2VVnBsvwJaWfZexJepzcimeaJ05ZoqrmkFyUQRREAYLfbU3wlREQ0UcLv6eH3+LFijCAimn4mKkYMPQfjBE1Vc67Jx7mTvbDbZ0rbHN1OAIDBpI17byfaznrCBltPPxaUFUrbvE4fug71YO41sy7IcyIar2TixCWdUBsYCK1QMmfOnBRfCRERTbSBgQFkZGSMqz3AGEFENB2NN0aEzwEwTtDUNUOTg69ctxE3VV8nbfvEovvQ4+rGFzL+FQCQrtCg6uovYdfh53G6/2TC7WZocvCJxZ9ES8fbEY95Rc5SvHnydelcRJNVInFCECfi45kpKhgM4syZM9Dr9RAEIdWXk1J2ux1z5szB6dOnYTAYUn05lwz2e+qw71PjYvS7KIoYGBjArFmzIJONfWYDxojz+HpJHfZ96rDvU+NC9/tExQiAcWIovl5SZ7x9bz1hw6m3zyDLZJSqzBb/n4XSfke3E3/+5hu48UsrkLc0J+F2jdWvwOfyxXzMT/3yY0lf52TDez51JlOcuKQTanSe3W5HRkYG+vv7+YZwEbHfU4d9nxrs96mJP7fUYd+nDvs+NdjvUxN/bqnDvk8N9nvqTKa+56IERERERERERERESWBCjYiIiIiIiIiIKAlMqBEAQK1W49vf/jbUanWqL+WSwn5PHfZ9arDfpyb+3FKHfZ867PvUYL9PTfy5pQ77PjXY76kzmfqec6gRERERERERERElgRVqRERERERERERESWBCjYiIiIiIiIiIKAlMqBERERERERERESWBCTUiIiIiIiIiIqIkKFJ9ATSxtm7dira2NpSWliIrKytqf0VFBQDAYrGgtrYWRUVFAACj0YiqqqqIYxM5hs5LtO9LS0uxefNmlJWVAQDq6+sBADU1NdKx7PuxMZvNaG5uBgD09vYiOzs7ol8B3vsXQiL9zvt+8mCcSB3GidRijEgdxomphXEiNRgjUo9xInWmbJwQaVqpqqoSAcT8V1JSIoqiKLa1tYlGo1Hs6+uT2tXU1Ii1tbXS94kcQ5ES6XtRFKP2VVVVRZyHfT82bW1tUX3U0tIiVlRURBzDe39iJdLvosj7fjJhnEgdxonUYYxIHcaJqYdxIjUYI1KLcSJ1pnKcYEJtmhl+U4XV1taKbW1t0jE1NTUR+/v6+sSh+dVEjqFIifR9+Li6ujqxrq4uYvvQ/ez75FVVVcXsz7KysohjeO9PrET6PXwc7/vJgXEidRgnUocxInUYJ6YexonUYIxILcaJ1JnKcYJzqE0z5eXlUdvMZjOMRiNMJhMAYMeOHVIJZJjRaAQAqcwykWMoUiJ9DwBFRUWoqqpCVVVVxPYw9v3YWK1W1NbWxtwexnt/4iXS7wDv+8mEcSJ1GCdShzEidRgnph7GidRgjEgtxonUmcpxggm1aSY8tn6ouro6adywzWaDzWaLeQMajUaYzeaEjqFoo/X9UDabDc3NzVF9yb4fu+rqatTX16OyshI2mw1AaC6K6upqALz3L5TR+n0o3veTA+NE6jBOpA5jROowTkw9jBOpwRiRWowTqTOV4wQTatPcxo0bI7K9Fosl7rFZWVno7e1N6Bga3fC+D2tqakJzczNWrFgBIPRpVPgFzr4fu7KyMtTW1qKxsRGZmZmorKxEWVmZ9EsI7/0LY7R+D+N9P3kxTqQO48TFwxiROowTUx/jRGowRlxcjBOpM5XjBFf5nMYsFgtsNptU5piIcEZ4vMdc6kbq+7q6OilzXlJSgurqalRWVqKtrW3U87LvR1ZRUYF9+/bBYrGgsbERALB9+/aEXgO898cukX7nfT85MU6kDuPExccYkTqME1MX40RqMEakBuNE6kzVOMEKtWmstrY2aiz+SG8G4THKiRxDI4vV92HDy1BLSkpgsVjQ3NzMvh8Hs9mMjRs3oqGhAS0tLdKnHKWlpQB4718oo/V7GO/7yYlxInUYJy4uxojUYZyY2hgnUoMx4uJjnEidqRwnmFCbxnbs2IGSkpKIbVlZWQBiZ2jDn4IkcgyNLFbfA6HS7eHjt8P9bbFY2PfjsH79ejQ0NEjf19TUoK2tDVarFfX19bz3L5DR+h3gfT+ZMU6kDuPExcUYkTqME1Mb40RqMEZcfIwTqTOV4wQTatNUc3NzzEn5jEYjjEZj3CxteXl5QsdQfPH6HghNrrh///6IbeF+NplM7PsxGvomOpTJZMLmzZvR0tLCe/8CSKTfAd73kxXjROowTlxcjBGpwzgxtTFOpAZjxMXHOJE6Uz1OMKE2TY00Kd+6deuixhmHjy8rK0v4GIptpL6vra2NmlwxXKLKvh87k8kUt9+NRqNULsx7f2Il2u+87ycnxonUYZy4uBgjUodxYmpjnEgNxoiLj3EidaZ8nBBpWqqpqRHj/Xjb2tpEk8kUdXxdXV1Sx1BsI/V9U1OT2NDQIH3f19cnmkymiG3s+7GpqakRa2trI7b19fWJFRUV0ve89ydeIv3O+35yYpxIHcaJi48xInUYJ6YuxonUYIxIDcaJ1JnKcUIQRVG8MKk6SqX6+nrU1tbGXfHCbDbjhRdewNVXXy1lbWtqapI+hqKN1vfNzc1oamoCEMqYV1dXR2XM2fdjU19fL5VkA0B2dvaY7mv2f3IS6Xfe95MP40TqME6kBmNE6jBOTE2ME6nBGJE6jBOpM1XjBBNqRERERERERERESeAcakRERERERERERElgQo2IiIiIiIiIiCgJTKgRERERERERERElgQk1IiIiIiIiIiKiJDChRkRERERERERElAQm1IiIiIiIiIiIiJLAhBoREREREREREVESmFAjIhrCZrOl+hJSwmKxpPoSiIgmPcYIIiIaCePEpYUJNSKij9hsNmzcuDHVl5ESZrMZzc3Nqb4MIqJJizGCMYKIaCSME5denGBCbYqyWCyorq5GZWUlqqursXXrVmzduhVA6IUc/nq6M5vNKC8vR2ZmZsQLuLm5GZmZmSNmyi0WCyorK5GZmTnm/rLZbDE/hUjk8aeD6fY8169fj9ra2pj7En3NNTY2orKyEoIgIDMzE9XV1THvkfAxRUVFk+L1WlFRgYaGhmnzs7zUMUaEMEak1nR7nowRjBHTCeNECONEak2358k4cQnGCZGmnNraWrGkpERsamqK2N7X1yfW1NSIZWVlYk1NTYquLjUARPRHS0uLWFJSIvb19Y3a1mQyibW1tWN63KamJrGlpSVqezKPP5VNp+dZV1cn1tXVxdw3ltdcSUmJaDKZRnzMsrKySdV3fX19YllZWaovg8aJMSIaY0RqTKfnyRjBGDGdME5EY5xIjen0PBknLs04wYTaFFNXVycajcYRXzgmk+mSC4JGozHqDSpRJSUlYw6CVVVVMYMgTT0lJSUxt4/1NVdXVycCiHt/9PX1jfm+u5CqqqrG/Fqi1GOMiI0xgsaLMSKEMWLqY5yIjXGCxotxIuRSixMc8jmFhMtEa2trYTQa4x53qY7bvtiam5tRX1+f6sugCdDc3IwVK1ZEbR/Pa66qqgoAUFdXF7NNfX29dMxkUllZGfeaaXJjjJhcGCOmD8aI8xgjpjbGicmFcWL6YJw471KLE4pUXwAlLjwee926dSMet27dOmzZsgVA6MW9ceNGmM1mtLW1obGxEfv27UN5ebn0AjSbzXjhhRdQVFQEIDSGu6amRjpfY2MjrFYrsrKyYLVa0dLSgsrKSpSVlY24L57GxkZs2bIFZrMZZWVlaGhogNFoRHNzMyorK5GVlYXa2lpUVFQAQESgaWlpQXV1NUpKSuKeP/zGtX//fjQ0NERci8ViQW1tLYqKimA0GpGVlRX3PCM9bnNzMxoaGgAAW7ZsgclkAhD6GY30+KP1tdlsxvr162GxWLB7925pDPq+ffuk8ycimT4e7XnGu3/KysriPs+RzjmW52g2m1FXV4fS0lLYbDYYjcaIANLc3Ayz2Qyj0ZjQPTJcQ0MDysvLo7aP5TU3VFVVFerr62MGlba2thEDa9jF+FkO7cuysjJUVlaOel00+TBGMEYwRjBGMEbQSBgnGCcYJxgnGCcmWKpL5ChxJpNJNBqNSbfr6+sTAUgloTU1NWJVVZUoiqFx+8PHOTc0NIgVFRWiKIpiW1ubdGxYbW2t2NTUNOK+RK9p+LE1NTViW1ub9H1dXV1E+WtbW5sIIOIYUYxdpj38/C0tLaLJZIootw2fb3i5bKKPixHKcIc//mh9PbxtVVVVxLUajUaxoaEh5mPFkkgfJ/I8R7p/Yj3PZPoukefY1NQUNX9ATU2NNEdBQ0NDVIm10WiMeryRmEymmMeP9TUX1tLSIgKImk+hqakpqVLoi/WzDIvXHzS5MUYwRjBGnH8OjBGMERSNcYJxgnHi/HNgnGCcmAgc8jmFWCyWET8FiWd45rq2tlbKcldXV0eVmFZUVKC5uRmNjY0wm81RK3WEP+0ZaV8i1xTOkA+VnZ0tfUITNnTFHZPJBKPRCLPZnNBjDFVZWYmNGzdGbDeZTFGPN97Hjff4o/X18Lbhf2ErVqyQPnlJ9PET6ePRnudI90+s/YmcM9wukecYq9+am5ul1W7Wr1+PzZs3R+xft25dwp/AAYDVao15H4z1NRdWUlKCkpKSqE+VmpqaRvzkdbiL9bMc2vaSW6FnGmCMYIxgjDh/fsYIxgiKxjjBOME4cf78jBOMExOBCbUpxGQywWq1jrn98LLVcBCLNd67rKwML7zwAsrKyrB//34UFRVh48aNaG5uhslkQllZ2Yj7ElFdXR1RVmo2m6OusaqqCi0tLQBCJc3hF3Sy/WCxWOI+11hv4hP1uGGJ9PVwV1999ajXOZrR+jiZ55lo2XMy5xztOcb7ubW0tKCmpgZmsxk2my3q2kpLS7F///6Erjd8nbGM9zUHhH4GZrNZ6gebzYbs7Owxnedi/SyNRuMlEwSnE8YIxohkMUYkhjEiEmPE1MU4wTiRLMaJxDBORLqU4gQTalNIWVkZbDZb3BfsUMM/pQAQlR0Pv0nEe3O1WCwwGo04fvy4NMdBeXk5ioqKpLHn8faVlpZCEISIf8M/VaioqIDRaJSutbm5OWYAbWxsRGlpqfTJwliy/OEXdDKBZCIeNyyRvr4QEunjRJ9nMs9/ovputJ9buF/Dn8yF/4XnAkjESK+n8b7mgOgJReNNIDraayZVP0uaOhgjGCOSxRgxOsYImk4YJxgnksU4MTrGiUsbFyWYQjZu3Ij6+nrs2LFjxBU9En1DDZd3hgNarP3hzHX4BRyeJHPLli2477774u4LZ7dHs27dOtTV1cUt7966dSvq6urQ1NQUt5w6EUOfayLG8rjhT0BiBfJE+vpCGamPJ6p/L9Q5w+0tFkvMc4W3lZWVjfmxRvrFaKJec0MnFO3t7Y35mIm8Zi7Wz9Jms13Qe5IuDMYIxoixYIwYGWNENMaIqYtxgnFiLBgnRsY4Ee1SihOsUJtCTCYT6urqsHHjxhHfzBsbGxOaf6CsrAzGj1b4GK65uRnl5eWwWCxRY6nr6uqksuN4+xJVXV0tLRkd65o3btyIurq6iBdkuOzUZrPFvPZYwvMbJFq6O9bHjZetT6SvL5SR+nii+vdCnTP8c4t1T4U/VYk1p0J4f6KMRmPM19REveaqq6sBhObeGF6anoyL9bMMr7ZFUwtjRAhjRHIYI0bHGBGJMWLqYpwIYZxIDuPE6BgnIl1KcYIJtSmmqqoKmzdvxm233RbzhV9fX5/UBIXbt2+PWqK3vr4eK1askLLow8tdLRaL9KY90r5ElJSUwGQyoaGhIW4We+ibT/jTBZvNFvVCjfUmNXRbXV1d1PU2NzfDYrGgt7d3TI9bUlIiBdZY8zYMPUcifT2SRMuFhxutjxPt39GubSLOGes51tXVYcuWLVHnDNu+fXvURKMWiyWpvhpp4syJeM2FfwbNzc0JT7Y70nku5M9yaFuaehgjGCOSxRgxOsaISIwRUxvjBONEshgnRsc4EelSihMc8jkF1dTUoKKiQnpDNxqN0sSEVVVVESWgzc3NUhn1+vXrUVZWFhEIKioqYDKZsHHjRmnOAiC0ckhYZWUltm7dKp3XZrOhpqYGjY2NcfclY/gb2FBNTU2oq6uDxWJBSUkJsrKy0NDQILUpKytDdXU1bDYbtmzZAqvVihUrVqCurk7aFn6e4ZVNNm7ciKuvvlpajcVkMqG+vh42m03qq9EeN/wcw2/AtiFlrWazOebjJ9LXw9tarVasW7cOW7Zswf79+6WAG2s1lbH08WjPs6SkBJWVlQCi7594z3MsP7ORnmNZWRl2794t9ZvJFFpxJhx4wv1aXV2N0tJSAKFP95IJNvfddx/2798fd4LNZF5z8dTW1ia1slI8F+JnOVT4Xh7LxLU0OTBGMEYwRjBGDMcYQUMxTjBOME4wTgzHODE2giiKYqovgogolSwWCzZu3Bi1lPSlqLGxEVarNaFPOomILgWMEecxRhARRWOcOO9SixNMqBERIbQyzu7duy+ZT1PiYT8QEUXje2MI+4GIKDa+P4Zcav3AOdSIiBB7XopLjdlsliZnJSKi8xgjGCOIiEbCOHFpxgkm1IiIEJoXIDs7O6mVpaaT8DwUseZCICK61DFGMEYQEY2EceLSjBNMqBERfaSmpgbNzc1jWgFpqquvr8f27dtTfRlERJMWYwRjBBHRSBgnLr04wTnUiIiIiIiIiIiIksAKNSIiIiIiIiIioiQwoUZERERERERERJQEJtSIiIiIiIiIiIiSwIQaERERERERERFREphQIyIiIiIiIiIiSgITakRERERERERERElgQo2IiIiIiIiIiCgJTKgRERERERERERElgQk1IiIiIiIiIiKiJPz/6Z5I1XdVQeAAAAAASUVORK5CYII=", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "total_samples_list = agg_df.loc[~(agg_df.index.get_level_values(\"algorithm\").isin((\"kl_empirical\", \"wasserstein_empirical\")))].index.get_level_values(\"num_total_samples\").unique().tolist()\n", + "# num_observations_list = agg_df.index.get_level_values(\"num_observations\").unique().tolist()\n", + "dgp_list = agg_df.index.get_level_values(\"dgp\").unique().tolist()\n", + "\n", + "nrows = len(dgp_list)\n", + "ncols = len(total_samples_list)\n", + "# ncols = len(num_observations_list)\n", + "trim_epsilon = 1.0\n", + "\n", + "fig, axes = plt.subplots(nrows=nrows, ncols=ncols, sharex=True, sharey=True, figsize=(5*ncols, nrows*4))\n", + "fig.subplots_adjust(wspace=0.05)\n", + "pd.options.mode.chained_assignment = None # default='warn'\n", + "for i, dgp in enumerate(dgp_list):\n", + " for j, total_samples in enumerate(total_samples_list):\n", + " # for j, num_observations in enumerate(num_observations_list):\n", + "\n", + " # NOTE empirical KL doesn't sample, so we plot it before filtering by total_samples\n", + " # mean_variance_plot(axes[j], agg_df.loc[\"kl_empirical\", dgp, :trim_epsilon, \"empirical\", :, :], **algorithm_inference_style(\"kl_empirical\", \"empirical\", label_inference=False), special_epsilons=[0.3], offset=0.1)\n", + " # mean_variance_plot(axes[j], agg_df.loc[\"kl_empirical\", dgp, :trim_epsilon, \"empirical\", num_observations, :], **algorithm_inference_style(\"kl_empirical\", \"empirical\", label_inference=False), special_epsilons=[0.3], offset=0.1)\n", + " # mean_variance_plot(axes[j], agg_df.loc[\"wasserstein_empirical\", dgp, :, \"empirical\", :, :], **algorithm_inference_style(\"wasserstein_empirical\", \"empirical\", label_inference=False), special_epsilons=[0.3], offset=0.1)\n", + "\n", + "\n", + " # filter by the total samples and DGP\n", + " axis_df = agg_df.loc[:, dgp, :, :, total_samples, :]\n", + " # axis_df = agg_df.loc[:, dgp, :, :, :, num_observations]\n", + " axis_df.loc[:, \"is_pareto_front\"] = is_minimise_pareto_front(axis_df[\"out_of_sample_var\"].values, axis_df[\"out_of_sample_mean\"].values)\n", + " for algorithm, inference in set(zip(axis_df.index.get_level_values(\"algorithm\"), axis_df.index.get_level_values(\"inference\"))):\n", + " df = axis_df.loc[algorithm, :trim_epsilon, :, :]\n", + "\n", + " if algorithm == \"kl_bdro\":\n", + " print(\"All BDRO points are Pareto dominated for M =\", total_samples, \"?\", not df[\"is_pareto_front\"].any())\n", + "\n", + " min_vf_mean_row = df.loc[df[\"out_of_sample_mean\"] == df[\"out_of_sample_mean\"].min()].iloc[0]\n", + " min_vf_var_row = df.loc[df[\"out_of_sample_var\"] == df[\"out_of_sample_var\"].min()].iloc[0]\n", + "\n", + "\n", + " weighted_VF = 0.5 * (df[\"out_of_sample_mean\"] + df[\"out_of_sample_std\"])\n", + " vf_comprimise_row = df.loc[weighted_VF == weighted_VF.min()].iloc[0]\n", + "\n", + " special_epsilons = [min_vf_mean_row.name[1], vf_comprimise_row.name[1], min_vf_var_row.name[1]]\n", + "\n", + " for k in range(j, len(total_samples_list)):\n", + " # for k in range(j, len(num_observations_list)):\n", + " alpha = 1.0\n", + " is_labelled=True\n", + " if j != k:\n", + " alpha = 0.2\n", + " is_labelled = False\n", + " if j == 0:\n", + " offset = 0.2\n", + " else:\n", + " offset = 0.1\n", + " mean_variance_plot(axes[k], df, **algorithm_inference_style(algorithm, inference, label_inference=False), offset=offset, is_labelled=is_labelled, alpha=alpha, special_epsilons=special_epsilons, add_end_epsilons=False)\n", + " \n", + " axes[j].scatter(min_vf_mean_row[\"out_of_sample_var\"], min_vf_mean_row[\"out_of_sample_mean\"], marker=\"x\", color=AlgorithmColor[algorithm].value, label=\"$\\min$ CV-mean\", s=100)\n", + " axes[j].scatter(min_vf_var_row[\"out_of_sample_var\"], min_vf_var_row[\"out_of_sample_mean\"], marker=\"*\", color=AlgorithmColor[algorithm].value, label=\"$\\min$ CV-var\", s=100)\n", + " axes[j].scatter(vf_comprimise_row[\"out_of_sample_var\"], vf_comprimise_row[\"out_of_sample_mean\"], marker=\"^\", color=AlgorithmColor[algorithm].value, label=\"$\\min$ CV-mean-std\", s=100)\n", + "\n", + " if j == 0:\n", + " axes[j].set_ylabel(\"Cross-validation mean (CV-mean)\")\n", + " if j==2:\n", + " handles, labels = axes[j].get_legend_handles_labels()\n", + " algorithm_order = [AlgorithmName.kl_dro_bas.value, AlgorithmName.kl_pp.value]\n", + " # algorithm_order = [AlgorithmName.kl_dro_bas.value, AlgorithmName.kl_pp.value, AlgorithmName.kl_bdro.value, AlgorithmName.wasserstein_empirical.value]\n", + " # order = [list(labels).index(a) for a in algorithm_order]\n", + " # axes[j].legend([handles[idx] for idx in order],[labels[idx] for idx in order])\n", + " axes[j].legend()\n", + " axes[j].set_title(\"$M$\" + f\"={total_samples} with {NiceNameDGP[filter_dgp]}\")\n", + " # axes[j].set_title(\"$n$\" + f\"={num_observations} with {NiceNameDGP[filter_dgp]}\")\n", + " axes[j].set_xlabel(\"Cross-validation variance (CV-var)\") \n", + "\n", + "\n", + "fig.savefig(f\"/Users/patrick/Experiments/misdro/2025_03_28_cross_validation/splits_newsvendor_{filter_dgp}_100_observations.pdf\", bbox_inches=\"tight\")" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "**Plot the OOS mean and variance when we solve with $\\epsilon_\\text{CV}^j$ for each iteration $j$**" + ] + }, + { + "cell_type": "code", + "execution_count": 16, + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/Users/patrick/Projects/mis-dro-code/mis_dro/results.py:66: FutureWarning: The provided callable is currently using SeriesGroupBy.mean. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"mean\" instead.\n", + " agg_df = gb.agg(\n", + "/Users/patrick/Projects/mis-dro-code/mis_dro/results.py:66: FutureWarning: The provided callable is currently using SeriesGroupBy.std. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"std\" instead.\n", + " agg_df = gb.agg(\n" + ] + } + ], + "source": [ + "# get the experiment that contains Newsvendor results for Normal DGP with 100 observations (instead of the usual 20)\n", + "experiment_100_dir = Path(f\"/Users/patrick/experiments/misdro/2025_03_28_cross_validation/newsvendor_100\")\n", + "all_results_100_df = pd.read_csv(experiment_100_dir / \"results.csv\", index_col=[\"uuid\", \"replication\"])\n", + "results_100_df = preprocess_results_df(all_results_100_df, filter_dgp)\n", + "agg_100_df = get_agg_df(results_100_df, [\"algorithm\", \"dgp\", \"epsilon\", \"inference\", \"num_total_samples\", \"num_observations\"])\n" + ] + }, + { + "cell_type": "code", + "execution_count": 17, + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "total_samples_list = agg_100_df.loc[~(agg_100_df.index.get_level_values(\"algorithm\").isin((\"kl_empirical\", \"wasserstein_empirical\")))].index.get_level_values(\"num_total_samples\").unique().tolist()\n", + "# num_observations_list = agg_df.index.get_level_values(\"num_observations\").unique().tolist()\n", + "dgp_list = agg_100_df.index.get_level_values(\"dgp\").unique().tolist()\n", + "\n", + "nrows = len(dgp_list)\n", + "ncols = len(total_samples_list)\n", + "# ncols = len(num_observations_list)\n", + "trim_epsilon = 1.0\n", + "\n", + "fig, axes = plt.subplots(nrows=nrows, ncols=ncols, sharex=True, sharey=True, figsize=(5*ncols, nrows*4))\n", + "fig.subplots_adjust(wspace=0.05)\n", + "pd.options.mode.chained_assignment = None # default='warn'\n", + "for i, dgp in enumerate(dgp_list):\n", + " for j, total_samples in enumerate(total_samples_list):\n", + "\n", + " # filter by the total samples and DGP\n", + " axis_df = agg_100_df.loc[:, dgp, :, :, total_samples, :]\n", + " # axis_df = agg_100_df.loc[:, dgp, :, :, :, num_observations]\n", + " axis_df.loc[:, \"is_pareto_front\"] = is_minimise_pareto_front(axis_df[\"out_of_sample_var\"].values, axis_df[\"out_of_sample_mean\"].values)\n", + " for algorithm, inference in set(zip(axis_df.index.get_level_values(\"algorithm\"), axis_df.index.get_level_values(\"inference\"))):\n", + " df = axis_df.loc[algorithm, :trim_epsilon, :, :]\n", + "\n", + " if algorithm == \"kl_bdro\":\n", + " print(\"All BDRO points are Pareto dominated for M =\", total_samples, \"?\", not df[\"is_pareto_front\"].any())\n", + "\n", + "\n", + " for k in range(j, len(total_samples_list)):\n", + " # for k in range(j, len(num_observations_list)):\n", + " alpha = 1.0\n", + " is_labelled=True\n", + " if j != k:\n", + " alpha = 0.2\n", + " is_labelled = False\n", + " mean_variance_plot(axes[k], df, **algorithm_inference_style(algorithm, inference, label_inference=False), offset=0.0, is_labelled=is_labelled, alpha=alpha)\n", + " if j == 0:\n", + " handles, labels = axes[j].get_legend_handles_labels()\n", + " algorithm_order = [AlgorithmName.kl_dro_bas.value, AlgorithmName.kl_pp.value]\n", + " # algorithm_order = [AlgorithmName.kl_dro_bas.value, AlgorithmName.kl_pp.value, AlgorithmName.kl_bdro.value, AlgorithmName.wasserstein_empirical.value]\n", + " order = [list(labels).index(a) for a in algorithm_order]\n", + " axes[j].legend([handles[idx] for idx in order],[labels[idx] for idx in order])\n", + " axes[j].set_ylabel(\"Out-of-sample mean, $m(\\epsilon)$\")\n", + " axes[j].set_title(\"$M$\" + f\"={total_samples} with {NiceNameDGP[filter_dgp]}\")\n", + " # axes[j].set_title(\"$n$\" + f\"={num_observations} with {NiceNameDGP[filter_dgp]}\")\n", + " axes[j].set_xlabel(\"Out-of-sample variance, $v(\\epsilon)$\")\n", + "\n", + " # plot where the CV result is on the curve\n", + " for idx, row in cv_agg_df.loc[cv_agg_df.index.get_level_values(\"num_total_samples\") == total_samples].iterrows():\n", + " algorithm = idx[0]\n", + " axes[j].scatter(row[\"out_of_sample_var\"], row[\"out_of_sample_mean\"], marker=\"x\", color=AlgorithmColor[algorithm].value, label=AlgorithmName[algorithm].value, s=200)\n", + "\n", + "\n", + "fig.savefig(f\"/Users/patrick/Experiments/misdro/2025_03_28_cross_validation/cross_validation_newsvendor_{filter_dgp}_100_observations.pdf\", bbox_inches=\"tight\")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [] + } + ], + "metadata": { + "kernelspec": { + "display_name": "mis-dro", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.11.6" + }, + "orig_nbformat": 4 + }, + "nbformat": 4, + "nbformat_minor": 2 +} diff --git a/notebooks/cv_portfolio.ipynb b/notebooks/cv_portfolio.ipynb new file mode 100644 index 0000000..1354f66 --- /dev/null +++ b/notebooks/cv_portfolio.ipynb @@ -0,0 +1,616 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# Portfolio problem" + ] + }, + { + "cell_type": "code", + "execution_count": 1, + "metadata": {}, + "outputs": [], + "source": [ + "import math\n", + "from pathlib import Path\n", + "import cvxpy as cp\n", + "import numpy as np\n", + "import pandas as pd\n", + "import scipy as sp\n", + "import matplotlib.pyplot as plt\n", + "\n", + "from mis_dro.dataset import get_portfolio_returns_df, get_num_time_windows, portfolio_dataset\n", + "from mis_dro.experiments import ExperimentName\n", + "from mis_dro.constants import IN_SAMPLE_TIME_WINDOW, OUT_OF_SAMPLE_TIME_WINDOW\n", + "from mis_dro.plot import *\n", + "from mis_dro.portfolio import portfolio_objective_cvxpy\n", + "from mis_dro.preprocessing import normalise_by_dimension\n", + "from mis_dro.results import preprocess_results_df, is_maximise_pareto_front, is_minimise_pareto_front, get_agg_df, get_result_df_list, convert_str_to_float_list\n", + "\n", + "# from matplotlib import rc\n", + "# rc('font', **{'family': 'serif', 'serif': ['Computer Modern'], \"size\":14})\n", + "# rc('text', usetex=True)\n" + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "metadata": {}, + "outputs": [], + "source": [ + "mmc2_dir = Path(\"/\", \"Users\", \"patrick\", \"Datasets\", \"misdro\", \"mmc2\")\n", + "# mmc2_dir = Path(\"/dcs/pg20/u1508153/datasets/misdro/mmc2\")\n", + "# assert mmc2_dir.exists()\n", + "dgp = \"DowJones\"\n", + "returns_df = get_portfolio_returns_df(mmc2_dir, dgp)\n", + "index_returns_df = pd.read_excel(mmc2_dir / \"Datasets\" / dgp / f\"{dgp}.xlsx\", sheet_name=\"Index_Returns\", header=None)" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "metadata": {}, + "outputs": [ + { + "data": { + "text/plain": [ + "array(['kl_dro_bas', 'kl_pp'], dtype=object)" + ] + }, + "execution_count": 3, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "experiment_name = ExperimentName.kl_portfolio\n", + "# kl_experiment_dir = Path(\"/dcs/large/u1508153/misdro/2025_01_21_paper_bas_experiments/kl_portfolio\")\n", + "kl_experiment_dir = Path(f\"/Users/patrick/experiments/misdro/2025_01_21_paper_bas_experiments/{experiment_name.value}\")\n", + "assert kl_experiment_dir.exists()\n", + "kl_results_df = pd.read_csv(kl_experiment_dir / \"results.csv\", index_col=[\"uuid\", \"replication\"])\n", + "\n", + "# mmd_experiment_dir = Path(\"/dcs/large/u1508153/misdro/mmd_portfolio_debug/2025_01_17_normalise\")\n", + "# experiment_dir = Path(f\"/Users/patrick/experiments/misdro/{experiment_name.value}\")\n", + "# assert mmd_experiment_dir.exists()\n", + "# mmd_results_df = pd.read_csv(mmd_experiment_dir / \"results.csv\", index_col=[\"uuid\", \"replication\"])\n", + "\n", + "all_results_df = kl_results_df.loc[kl_results_df[\"algorithm\"].isin([\"kl_dro_bas\", \"kl_pp\"])]\n", + "# all_results_df = pd.concat([kl_results_df, mmd_results_df])\n", + "\n", + "all_results_df[\"algorithm\"].unique()" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "metadata": {}, + "outputs": [], + "source": [ + "# NOTE filter by DGP\n", + "filter_dgp = \"DowJones\"\n", + "dataset = \"portfolio\"\n", + "results_df = preprocess_results_df(all_results_df, filter_dgp, dataset=dataset)\n", + "\n", + "# NOTE filter out when log_partition_constant G is less than epsilon\n", + "# results_df = results_df.loc[results_df[\"log_partition_constant\"] < results_df[\"epsilon\"]]\n", + "\n", + "# NOTE we always want the same number of test observations and replications\n", + "assert np.isclose(results_df[\"num_test_observations\"].var(), 0)\n", + "num_test_observations = results_df[\"num_test_observations\"].unique()[0]\n", + "assert np.isclose(results_df[\"num_replications\"].var(), 0)\n", + "num_replications = results_df[\"num_replications\"].unique()[0]\n", + "\n", + "# NOTE if you are running PP with multiple num_likelihood_samples then you will need the below line\n", + "results_df = results_df.loc[((results_df[\"algorithm\"] == \"kl_pp\") & (results_df[\"num_likelihood_samples\"] == 3600)) | (results_df[\"algorithm\"] != \"kl_pp\")]" + ] + }, + { + "cell_type": "code", + "execution_count": 7, + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/Users/patrick/Projects/mis-dro-code/mis_dro/results.py:66: FutureWarning: The provided callable is currently using SeriesGroupBy.mean. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"mean\" instead.\n", + " agg_df = gb.agg(\n", + "/Users/patrick/Projects/mis-dro-code/mis_dro/results.py:66: FutureWarning: The provided callable is currently using SeriesGroupBy.std. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"std\" instead.\n", + " agg_df = gb.agg(\n" + ] + }, + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
out_of_sample_meanout_of_sample_varsum_of_in_group_varvar_of_in_group_meanmean_solve_timestd_solve_timemean_sample_timestd_sample_time
algorithm
kl_dro_bas0.0053260.0018170.0016880.0001290.0087040.0029970.0002870.000244
kl_pp0.0037030.0012840.0012130.0000720.2180930.0290470.0015180.001638
\n", + "
" + ], + "text/plain": [ + " out_of_sample_mean out_of_sample_var sum_of_in_group_var \\\n", + "algorithm \n", + "kl_dro_bas 0.005326 0.001817 0.001688 \n", + "kl_pp 0.003703 0.001284 0.001213 \n", + "\n", + " var_of_in_group_mean mean_solve_time std_solve_time \\\n", + "algorithm \n", + "kl_dro_bas 0.000129 0.008704 0.002997 \n", + "kl_pp 0.000072 0.218093 0.029047 \n", + "\n", + " mean_sample_time std_sample_time \n", + "algorithm \n", + "kl_dro_bas 0.000287 0.000244 \n", + "kl_pp 0.001518 0.001638 " + ] + }, + "execution_count": 7, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "# cv_kl_portfolio_dir = Path(\"/dcs/large/u1508153/misdro/cv_kl_portfolio\")\n", + "cv_kl_portfolio_dir = Path(\"/Users/patrick/Experiments/misdro/2025_03_28_cross_validation/cv_portfolio_different_replications\")\n", + "cv_kl_portfolio_df = pd.read_csv(cv_kl_portfolio_dir / \"results.csv\", index_col=[\"uuid\", \"replication\"])\n", + "cv_df = cv_kl_portfolio_df.loc[(cv_kl_portfolio_df[\"use_cv_epsilon\"]) & (cv_kl_portfolio_df[\"algorithm\"].isin([\"kl_dro_bas\", \"kl_pp\"]))]\n", + "cv_df = preprocess_results_df(cv_df, dgp, dataset=\"portfolio\")\n", + "cv_agg_df = get_agg_df(cv_df, [\"algorithm\"])\n", + "cv_agg_df\n" + ] + }, + { + "cell_type": "code", + "execution_count": 8, + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/Users/patrick/Projects/mis-dro-code/mis_dro/results.py:66: FutureWarning: The provided callable is currently using SeriesGroupBy.mean. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"mean\" instead.\n", + " agg_df = gb.agg(\n", + "/Users/patrick/Projects/mis-dro-code/mis_dro/results.py:66: FutureWarning: The provided callable is currently using SeriesGroupBy.std. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"std\" instead.\n", + " agg_df = gb.agg(\n" + ] + } + ], + "source": [ + "agg_df = get_agg_df(results_df, [\"algorithm\", \"epsilon\", \"inference\"])\n" + ] + }, + { + "cell_type": "code", + "execution_count": 10, + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/var/folders/38/gm_d4yxj7bx11rp4p930zwnr0000gn/T/ipykernel_79909/315046132.py:33: UserWarning: FigureCanvasAgg is non-interactive, and thus cannot be shown\n", + " fig.show()\n" + ] + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "fig, ax = plt.subplots(figsize=(8,6))\n", + "df = agg_df\n", + "df.loc[:, \"is_pareto_front\"] = is_maximise_pareto_front(df[\"out_of_sample_var\"].values, df[\"out_of_sample_mean\"].values)\n", + "for (algorithm, inference), group_df in df.groupby([\"algorithm\", \"inference\"]):\n", + " if algorithm == \"kl_dro_bas\":\n", + " local_log_partition_constant = True\n", + " offset=0.00003\n", + " special_epsilons = [0.00001, 0.001, 0.01, 1.0]\n", + " else:\n", + " local_log_partition_constant = False\n", + " offset=0.00003\n", + " special_epsilons = [0.00001, 0.0005, 0.005, 1.0]\n", + " mean_variance_plot(\n", + " ax, group_df, is_labelled=True, offset=offset, alpha=1.0,\n", + " **algorithm_inference_style(algorithm, inference, label_inference=False),\n", + " special_epsilons=special_epsilons,\n", + " add_log_partition_function=local_log_partition_constant, minimise=False\n", + " )\n", + " # log_partition_constant=local_log_partition_constant\n", + "\n", + "for algorithm, row in cv_agg_df.iterrows():\n", + " ax.scatter(row[\"out_of_sample_var\"], row[\"out_of_sample_mean\"], marker=\"x\", color=AlgorithmColor[algorithm].value, label=AlgorithmName[algorithm].value)\n", + "\n", + "ax.legend(fontsize=12)\n", + "ax.set_xlabel(\"Out-of-sample variance\")\n", + "ax.set_ylabel(\"Out-of-sample mean\")\n", + "# ax.set_xlim(agg_df[\"out_of_sample_var\"].min(), agg_df[\"out_of_sample_var\"].max())\n", + "# ax.set_ylim(agg_df[\"out_of_sample_mean\"].min(), agg_df[\"out_of_sample_mean\"].min())\n", + "handles, labels = ax.get_legend_handles_labels()\n", + "algorithm_order = [AlgorithmName.kl_dro_bas.value, AlgorithmName.kl_pp.value]\n", + "order = [list(labels).index(a) for a in algorithm_order]\n", + "ax.legend([handles[idx] for idx in order],[labels[idx] for idx in order])\n", + "fig.show()\n", + "# fig.savefig(f\"/Users/patrick/Experiments/misdro/2025_01_21_paper_bas_experiments/{experiment_name}_mean_var_{filter_dgp}.pdf\", bbox_inches=\"tight\")\n" + ] + }, + { + "cell_type": "code", + "execution_count": 13, + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/var/folders/38/gm_d4yxj7bx11rp4p930zwnr0000gn/T/ipykernel_79909/430931101.py:97: UserWarning: FigureCanvasAgg is non-interactive, and thus cannot be shown\n", + " fig.show()\n" + ] + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "# fig, ax = plt.subplots(figsize=(8,6))\n", + "fig, ax = plt.subplots(figsize=(16,9))\n", + "\n", + "num_time_windows = math.floor((len(returns_df) - IN_SAMPLE_TIME_WINDOW) / OUT_OF_SAMPLE_TIME_WINDOW)\n", + "num_stocks = int(len(returns_df.columns))\n", + "last_week = len(index_returns_df.iloc[IN_SAMPLE_TIME_WINDOW:])\n", + "algorithm_labelled = {algorithm: False for algorithm in results_df[\"algorithm\"].unique()}\n", + "kl_epsilon_list = [0.00001, 0.001, 1.0]\n", + "mmd_epsilon_list = [0.0001, 0.01, 0.2]\n", + "line_styles = [(0, (1, 5)), \"dotted\", \"solid\"]\n", + "\n", + "# kl_algorithms = (\"kl_dro_bas\", \"kl_pp\", \"kl_bdro\", \"kl_empirical\")\n", + "kl_algorithms = (\"kl_dro_bas\", \"kl_pp\", \"kl_bdro\")\n", + "\n", + "oos_week_numbers = [IN_SAMPLE_TIME_WINDOW + i for i in range(OUT_OF_SAMPLE_TIME_WINDOW * num_time_windows)]\n", + "# for (algorithm, epsilon), group_df in results_df.groupby([\"algorithm\", \"epsilon\"]):\n", + "# # assert OUT_OF_SAMPLE_TIME_WINDOW * num_time_windows == len(group_df) * len(group_df[\"out_of_sample_cost\"].values[0])\n", + "# is_mmd_experiment = algorithm in (\"dro_bas_mmd\", \"empirical_mmd\")\n", + "# is_kl_experiment = algorithm in kl_algorithms\n", + "# if (is_kl_experiment and epsilon in kl_epsilon_list) or (is_mmd_experiment and epsilon in mmd_epsilon_list):\n", + "# cost_list = []\n", + "# for i in range(num_replications):\n", + "# cost_list += group_df[\"out_of_sample_cost\"].values[i]\n", + "# dro_returns = np.array(cost_list)\n", + "# # if not algorithm_labelled[algorithm]:\n", + "# # label = AlgorithmName[algorithm].value\n", + "# # algorithm_labelled[algorithm] = True\n", + "# # else:\n", + "# # label='_nolegend_'\n", + "# if algorithm == \"kl_dro_bas\":\n", + "# label = AlgorithmName[algorithm].value + \" ($G +\" + str(epsilon) + \"$)\"\n", + "# else:\n", + "# label = AlgorithmName[algorithm].value + f\" (${epsilon}$)\"\n", + "# is_robas_experiment = algorithm in (\"dro_bas_mmd\", \"empirical_mmd\")\n", + "# idx = mmd_epsilon_list.index(epsilon) if is_robas_experiment else kl_epsilon_list.index(epsilon)\n", + "# ax.plot(\n", + "# oos_week_numbers,\n", + "# np.cumsum(dro_returns),\n", + "# label=label,\n", + "# alpha=0.8,\n", + "# color=AlgorithmColor[algorithm].value,\n", + "# linestyle=line_styles[idx],\n", + "# )\n", + " # ax.text(last_week + IN_SAMPLE_TIME_WINDOW, np.cumsum(dro_returns)[last_week-4], epsilon, color=AlgorithmColor[algorithm].value)\n", + " # if algorithm == \"kl_dro_bas\" and epsilon in kl_epsilon_list:\n", + " # ax.text(last_week + IN_SAMPLE_TIME_WINDOW, np.cumsum(dro_returns)[last_week-4]-0.1, np.round(log_partition_constant + epsilon, 5), color=AlgorithmColor[algorithm].value)\n", + " # elif algorithm == \"kl_bdro\" and epsilon in kl_epsilon_list:\n", + " # ax.text(last_week + IN_SAMPLE_TIME_WINDOW, np.cumsum(dro_returns)[last_week-4]-0.1, epsilon, color=AlgorithmColor[algorithm].value)\n", + " # elif algorithm == \"dro_bas_mmd\" and epsilon in mmd_epsilon_list:\n", + " # ax.text(last_week + IN_SAMPLE_TIME_WINDOW, np.cumsum(dro_returns)[last_week-4], epsilon, color=AlgorithmColor[algorithm].value)\n", + " # elif algorithm == \"empirical_mmd\" and epsilon in mmd_epsilon_list:\n", + " # ax.text(last_week + IN_SAMPLE_TIME_WINDOW, np.cumsum(dro_returns)[last_week-4], epsilon, color=AlgorithmColor[algorithm].value)\n", + "\n", + "for algorithm, group_df in cv_df.groupby(\"algorithm\"):\n", + " cost_list = []\n", + " for i in range(num_replications):\n", + " cost_list += group_df[\"out_of_sample_cost\"].values[i]\n", + " dro_returns = np.array(cost_list)\n", + " label=f\"{AlgorithmName[algorithm].value} (10-fold CV)\"\n", + " ax.plot(\n", + " oos_week_numbers,\n", + " np.cumsum(dro_returns),\n", + " label=label,\n", + " alpha=0.8,\n", + " color=AlgorithmColor[algorithm].value,\n", + " )\n", + "\n", + "CB_color_cycle = ['#4daf4a',\n", + " '#f781bf', '#a65628', '#984ea3',\n", + " '#999999', '#e41a1c', '#dede00']\n", + "ax.plot(oos_week_numbers, np.cumsum(index_returns_df.iloc[IN_SAMPLE_TIME_WINDOW:IN_SAMPLE_TIME_WINDOW+OUT_OF_SAMPLE_TIME_WINDOW * num_time_windows].values), label=f\"{dgp} Index\", color=\"#a65628\")\n", + "# for baseline_model in [\"CZeSD\", \"KP_SSD\", \"L_SSD\", \"LR_ASSD\", \"MeanVar\", \"RMZ_SSD\"]:\n", + "\n", + "\n", + "\n", + "for i, baseline_model in enumerate([\"MeanVar\"]):\n", + " baseline_model_label = baseline_model if baseline_model != \"MeanVar\" else \"Markowitz\"\n", + " baseline_portfolio_txt = mmc2_dir / \"Solutions\" / dgp / f\"OptPortfolios_{baseline_model}_{dgp}.txt\"\n", + " baseline_portfolio = np.loadtxt(baseline_portfolio_txt).T[:num_time_windows, :]\n", + " baseline_returns_txt = mmc2_dir / \"Solutions\" / dgp / f\"OutofSamplePortReturns_{baseline_model}_{dgp}_List.txt\" \n", + " baseline_returns = np.loadtxt(baseline_returns_txt)\n", + " ax.plot(oos_week_numbers, np.cumsum(baseline_returns[:OUT_OF_SAMPLE_TIME_WINDOW * num_time_windows]), label=baseline_model_label, color=CB_color_cycle[i+1])\n", + "\n", + "ax.set_xlabel(\"Week number\")\n", + "ax.set_ylabel(\"Out-of-sample cumulative return\")\n", + "# ax.set_xlim(2000)\n", + "# ax.set_xlim(0, 1520)\n", + "# ax.set_ylim(-0.1, 9.5)\n", + "ax.vlines(52, -0.3, 6.2, color=\"black\", linestyle=\":\")\n", + "# ax.set_title(dgp + r\" cumulative return vs Benchmarks\")\n", + "# handles, labels = ax.get_legend_handles_labels()\n", + "# algorithm_order = [AlgorithmName.kl_dro_bas.value, AlgorithmName.kl_pp.value, AlgorithmName.kl_bdro.value, AlgorithmName.kl_empirical.value]\n", + "# order = [list(labels).index(a) for a in algorithm_order]\n", + "# ax.legend([handles[idx] for idx in order],[labels[idx] for idx in order])\n", + "ax.legend() \n", + "plt.savefig(f\"/Users/patrick/Experiments/misdro/2025_03_28_cross_validation/portfolio_cross_validation_cum_returns_{filter_dgp}.pdf\", bbox_inches=\"tight\")\n", + "fig.show()" + ] + }, + { + "cell_type": "code", + "execution_count": 18, + "metadata": {}, + "outputs": [ + { + "ename": "NameError", + "evalue": "name 'is_mmd_experiment' is not defined", + "output_type": "error", + "traceback": [ + "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", + "\u001b[0;31mNameError\u001b[0m Traceback (most recent call last)", + "Cell \u001b[0;32mIn[18], line 6\u001b[0m\n\u001b[1;32m 4\u001b[0m row_list \u001b[38;5;241m=\u001b[39m []\n\u001b[1;32m 5\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m (algorithm, epsilon), group_df \u001b[38;5;129;01min\u001b[39;00m results_df\u001b[38;5;241m.\u001b[39mloc[results_df[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124malgorithm\u001b[39m\u001b[38;5;124m\"\u001b[39m]\u001b[38;5;241m.\u001b[39misin(kl_algorithms)]\u001b[38;5;241m.\u001b[39mgroupby([\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124malgorithm\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mepsilon\u001b[39m\u001b[38;5;124m\"\u001b[39m]):\n\u001b[0;32m----> 6\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m (algorithm \u001b[38;5;129;01min\u001b[39;00m kl_algorithms \u001b[38;5;129;01mand\u001b[39;00m epsilon \u001b[38;5;129;01min\u001b[39;00m kl_epsilon_list) \u001b[38;5;129;01mor\u001b[39;00m (\u001b[43mis_mmd_experiment\u001b[49m \u001b[38;5;129;01mand\u001b[39;00m epsilon \u001b[38;5;129;01min\u001b[39;00m mmd_epsilon_list):\n\u001b[1;32m 7\u001b[0m cost_list \u001b[38;5;241m=\u001b[39m []\n\u001b[1;32m 8\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m i \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28mrange\u001b[39m(num_replications):\n", + "\u001b[0;31mNameError\u001b[0m: name 'is_mmd_experiment' is not defined" + ] + } + ], + "source": [ + "# is_mmd_experiment = algorithm in (\"dro_bas_mmd\", \"empirical_mmd\")\n", + "# is_kl_experiment = algorithm in (\"kl_dro_bas\", \"kl_pp\", \"kl_bdro\", \"kl_empirical\")\n", + "# if (is_kl_experiment and epsilon in kl_epsilon_list) or (is_mmd_experiment and epsilon in mmd_epsilon_list):\n", + "row_list = []\n", + "for (algorithm, epsilon), group_df in results_df.loc[results_df[\"algorithm\"].isin(kl_algorithms)].groupby([\"algorithm\", \"epsilon\"]):\n", + " if (algorithm in kl_algorithms and epsilon in kl_epsilon_list) or (is_mmd_experiment and epsilon in mmd_epsilon_list):\n", + " cost_list = []\n", + " for i in range(num_replications):\n", + " cost_list += group_df[\"out_of_sample_cost\"].values[i]\n", + " row_list.append({\"cumulative_return\": np.sum(cost_list), \"epsilon\": epsilon, \"algorithm\": algorithm})\n", + "cum_return_df = pd.DataFrame(row_list).set_index([\"algorithm\", \"epsilon\"])\n", + "cum_return_df = cum_return_df.reindex(pd.MultiIndex.from_arrays([np.repeat(kl_algorithms, len(kl_epsilon_list)), np.tile(kl_epsilon_list, len(kl_algorithms))], names=[\"algorithm\", \"epsilon\"]))\n", + "# solve_time_df.reindex(pd.MultiIndex.from_product(np.repeat([\"kl_dro_bas\", \"kl_pp\", \"kl_bdro\", \"kl_empirical\"], 3), solve_time_df.index.get_level_values(\"num_total_samples\"), names=['algorithm', 'num_total_samples']))\n", + "print(cum_return_df.unstack(level=\"algorithm\").to_latex())\n", + "# cum_return_df" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\\begin{tabular}{lll}\n", + "\\toprule\n", + " & str_solve_time & str_sample_time \\\\\n", + "algorithm & & \\\\\n", + "\\midrule\n", + "DRO-BAS$_{PE}$ & 0.01 (0.01) & 0.35 (0.30) \\\\\n", + "DRO-BAS$_{PP}$ & 1.34 (0.22) & 3.48 (0.91) \\\\\n", + "BDRO & 4.47 (0.62) & 44.45 (3.11) \\\\\n", + "Empirical KL & NaN & NaN \\\\\n", + "\\bottomrule\n", + "\\end{tabular}\n", + "\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/dcs/pg20/u1508153/projects/mis-dro-code/mis_dro/results.py:65: FutureWarning: The provided callable is currently using SeriesGroupBy.mean. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"mean\" instead.\n", + " agg_df = gb.agg(\n", + "/dcs/pg20/u1508153/projects/mis-dro-code/mis_dro/results.py:65: FutureWarning: The provided callable is currently using SeriesGroupBy.std. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"std\" instead.\n", + " agg_df = gb.agg(\n" + ] + } + ], + "source": [ + "solve_time_df = get_agg_df(results_df, [\"algorithm\"])[[\"mean_solve_time\", \"std_solve_time\", \"mean_sample_time\", \"std_sample_time\"]]\n", + "\n", + "decimal_places = 2\n", + "solve_time_df[\"str_solve_time\"] = solve_time_df[[\"mean_solve_time\", \"std_solve_time\"]].apply(\n", + " lambda x: \"{:.2f}\".format(np.round(x[\"mean_solve_time\"], decimal_places)) + \" (\" + \"{:.2f}\".format(np.round(x[\"std_solve_time\"], decimal_places)) + \")\", axis=1\n", + ")\n", + "solve_time_df[\"str_sample_time\"] = solve_time_df[[\"mean_sample_time\", \"std_sample_time\"]].apply(\n", + " lambda x: \"{:.2f}\".format(np.round(1000 * x[\"mean_sample_time\"], decimal_places)) + \" (\" + \"{:.2f}\".format(np.round(1000 * x[\"std_sample_time\"], decimal_places)) + \")\", axis=1\n", + ")\n", + "solve_time_df = solve_time_df.reindex([\"kl_dro_bas\", \"kl_pp\", \"kl_bdro\", \"kl_empirical\"])\n", + "solve_time_df.index = solve_time_df.index.map(lambda x: AlgorithmName[x].value)\n", + "print(solve_time_df[[\"str_solve_time\", \"str_sample_time\"]].to_latex())" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# bas_returns = []\n", + "# bdro_returns = []\n", + "# filter_epsilon = 0.1\n", + "# for (algorithm, epsilon), group_df in gb:\n", + "# # dro_returns = group_df[week_cols].values.flatten()\n", + "# if algorithm == \"kl_dro_bas\" and epsilon == filter_epsilon:\n", + "# assert OUT_OF_SAMPLE_TIME_WINDOW * num_time_windows == len(group_df) * len(group_df[\"out_of_sample_cost\"].values[0])\n", + "# for i in range(num_replications):\n", + "# bas_returns += group_df[\"out_of_sample_cost\"].values[i]\n", + "# if algorithm == \"kl_bdro\" and epsilon == filter_epsilon:\n", + "# assert OUT_OF_SAMPLE_TIME_WINDOW * num_time_windows == len(group_df) * len(group_df[\"out_of_sample_cost\"].values[0])\n", + "# for i in range(num_replications):\n", + "# bdro_returns += group_df[\"out_of_sample_cost\"].values[i]\n", + "# plt.hist(bas_returns, bins=50, label=\"BAS\", alpha=0.5)\n", + "# plt.hist(bdro_returns, bins=50, label=\"BDRO\", alpha=0.5)\n", + "# plt.legend()" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [ + "# from mis_dro.dataset import portfolio_dataset\n", + "\n", + "# # training_time_window_id = 75\n", + "# training_time_window_id = 79\n", + "# out_of_sample_time_window = IN_SAMPLE_TIME_WINDOW\n", + "# data, data_eval = portfolio_dataset(\"DowJones-crash\", training_time_window_id, mmc2_dir, out_of_sample_time_window=out_of_sample_time_window)\n", + "# dim = data.shape[1]\n", + "# sol = np.ones(dim) / float(dim)\n", + "# data.shape, data_eval.shape" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [ + { + "ename": "NameError", + "evalue": "name 'training_time_window_id' is not defined", + "output_type": "error", + "traceback": [ + "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", + "\u001b[0;31mNameError\u001b[0m Traceback (most recent call last)", + "Cell \u001b[0;32mIn[13], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m start_training_week \u001b[38;5;241m=\u001b[39m \u001b[43mtraining_time_window_id\u001b[49m \u001b[38;5;241m*\u001b[39m OUT_OF_SAMPLE_TIME_WINDOW \u001b[38;5;66;03m# inclusive\u001b[39;00m\n\u001b[1;32m 2\u001b[0m end_training_week \u001b[38;5;241m=\u001b[39m start_training_week \u001b[38;5;241m+\u001b[39m IN_SAMPLE_TIME_WINDOW \u001b[38;5;66;03m# not inclusive\u001b[39;00m\n\u001b[1;32m 3\u001b[0m test_start \u001b[38;5;241m=\u001b[39m end_training_week\n", + "\u001b[0;31mNameError\u001b[0m: name 'training_time_window_id' is not defined" + ] + } + ], + "source": [ + "start_training_week = training_time_window_id * OUT_OF_SAMPLE_TIME_WINDOW # inclusive\n", + "end_training_week = start_training_week + IN_SAMPLE_TIME_WINDOW # not inclusive\n", + "test_start = end_training_week\n", + "test_end = test_start + out_of_sample_time_window\n", + "total_training_returns = (data @ sol).sum()\n", + "\n", + "plt.plot(oos_week_numbers[start_training_week- IN_SAMPLE_TIME_WINDOW:test_end - IN_SAMPLE_TIME_WINDOW], np.cumsum(index_returns_df.iloc[start_training_week:test_end].values), label=f\"Out-of-sample {dgp} Index\", color=CB_color_cycle[0], linestyle=\"dotted\")\n", + "plt.plot(oos_week_numbers[start_training_week- IN_SAMPLE_TIME_WINDOW:test_start - IN_SAMPLE_TIME_WINDOW], np.cumsum(index_returns_df.iloc[start_training_week:test_start].values), label=f\"In-sample {dgp} Index\", color=CB_color_cycle[0])\n", + "\n", + "# plt.plot(oos_week_numbers[start_training_week - IN_SAMPLE_TIME_WINDOW:end_training_week - IN_SAMPLE_TIME_WINDOW], (data @ sol).cumsum())\n", + "# plt.plot(oos_week_numbers[test_start - IN_SAMPLE_TIME_WINDOW:test_end - IN_SAMPLE_TIME_WINDOW], (data_eval @ sol).cumsum())\n", + "plt.legend()\n" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [] + } + ], + "metadata": { + "kernelspec": { + "display_name": "mis-dro", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.11.6" + } + }, + "nbformat": 4, + "nbformat_minor": 2 +} diff --git a/notebooks/kde_epsilon_newsvendor.ipynb b/notebooks/kde_epsilon_newsvendor.ipynb new file mode 100644 index 0000000..1deeae5 --- /dev/null +++ b/notebooks/kde_epsilon_newsvendor.ipynb @@ -0,0 +1,923 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": 19, + "metadata": {}, + "outputs": [], + "source": [ + "import json\n", + "from pathlib import Path\n", + "import numpy as np\n", + "import pandas as pd\n", + "import scipy as sp\n", + "import matplotlib.pyplot as plt\n", + "\n", + "from mis_dro.results import preprocess_results_df, is_maximise_pareto_front, is_minimise_pareto_front, get_agg_df, get_result_df_list, convert_str_to_float_list\n", + "from mis_dro.plot import *\n", + "from mis_dro.experiments import ExperimentName\n", + "\n", + "# from matplotlib import rc\n", + "# rc('font', **{'family': 'serif', 'serif': ['Computer Modern'], \"size\":14})\n", + "# rc('text', usetex=True)" + ] + }, + { + "cell_type": "code", + "execution_count": 27, + "metadata": {}, + "outputs": [], + "source": [ + "experiment_name = ExperimentName.kl_newsvendor_1d\n", + "# experiment_dir = Path(f\"/Users/patrick/experiments/misdro/2025_01_21_paper_bas_experiments/{experiment_name.value}\")\n", + "experiment_dir = Path(f\"/dcs/large/u1508153/misdro/2025_01_21_paper_bas_experiments/{experiment_name.value}\")\n", + "# all_results_df = pd.read_csv(experiment_dir / \"results.csv\", index_col=[\"uuid\", \"replication\"])\n", + "\n", + "kde_experiment_dir = Path(f\"/dcs/large/u1508153/misdro/kde_epsilon_newsvendor_1d\")\n", + "kde_df = pd.read_csv(kde_experiment_dir / \"results.csv\", index_col=[\"uuid\", \"replication\"])\n" + ] + }, + { + "cell_type": "code", + "execution_count": 28, + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/dcs/pg20/u1508153/.conda/envs/mis-dro/lib/python3.11/site-packages/numpy/core/_methods.py:173: RuntimeWarning: invalid value encountered in subtract\n", + " x = asanyarray(arr - arrmean)\n" + ] + }, + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
out_of_sample_meanout_of_sample_var
algorithmdgp
kl_dro_basexponentialinfNaN
normal37.806960751.015391
truncated_normal32.207718534.362761
kl_ppnormal37.845161749.802710
truncated_normal32.310320529.791270
\n", + "
" + ], + "text/plain": [ + " out_of_sample_mean out_of_sample_var\n", + "algorithm dgp \n", + "kl_dro_bas exponential inf NaN\n", + " normal 37.806960 751.015391\n", + " truncated_normal 32.207718 534.362761\n", + "kl_pp normal 37.845161 749.802710\n", + " truncated_normal 32.310320 529.791270" + ] + }, + "execution_count": 28, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "filter_dgp = \"normal\" # filter by DGP\n", + "# kde_df.loc[kde_df[\"dgp\"] == filter_dgp]\n", + "num_test_observations = kde_df[\"num_test_observations\"].unique()[0]\n", + "kde_df[\"out_of_sample_cost\"] = kde_df[\"out_of_sample_cost\"].map(lambda x: convert_str_to_float_list(x, num_test_observations))\n", + "gb_cols=[\"algorithm\", \"dgp\"]\n", + "gb = kde_df.groupby(by=gb_cols)\n", + "agg_df = gb.agg(\n", + " out_of_sample_mean = pd.NamedAgg(column=\"out_of_sample_cost\", aggfunc=lambda x: np.mean(np.concatenate(x.values))),\n", + " out_of_sample_var = pd.NamedAgg(column=\"out_of_sample_cost\", aggfunc=lambda x: np.var(np.concatenate(x.values), ddof=1)),\n", + ")\n", + "agg_df\n" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
solutiondgp_timelikelihood_timeposterior_timesolve_timesetup_timelog_partition_constantout_of_sample_costalgorithmcontamination...num_likelihood_samplesnum_observationsnum_posterior_samplesnum_replicationsnum_test_observationsposteriornum_total_samplessample_timein_group_meanin_group_var
uuidreplication
\n", + "

0 rows × 31 columns

\n", + "
" + ], + "text/plain": [ + "Empty DataFrame\n", + "Columns: [solution, dgp_time, likelihood_time, posterior_time, solve_time, setup_time, log_partition_constant, out_of_sample_cost, algorithm, contamination, dataset, dgp, dim, epsilon, kde_epsilon, n_splits, ignore_dpp, inference, lengthscale, likelihood, njobs, num_likelihood_samples, num_observations, num_posterior_samples, num_replications, num_test_observations, posterior, num_total_samples, sample_time, in_group_mean, in_group_var]\n", + "Index: []\n", + "\n", + "[0 rows x 31 columns]" + ] + }, + "execution_count": 16, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [ + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
solutiondgp_timelikelihood_timeposterior_timesolve_timesetup_timelog_partition_constantout_of_sample_costalgorithmcontamination...num_posterior_samplesnum_replicationsnum_test_observationsposteriorkde_epsilonn_splitsnum_total_samplessample_timein_group_meanin_group_var
uuidreplication
10aee0fd-c2c4-448d-995a-b307828a920e0[26.504853109137507]0.0003260.0001190.0000810.042336NaN0.0[8.370599215733552, 97.27825277087481, 24.4703...kl_pp0.0...150050normal_gammaNaNNaN250.00020044.5541761543.100317
1[29.879991219010936]0.0002130.0000810.0000530.027507NaN0.0[14.395708241482502, 22.90806081601392, 64.485...kl_pp0.0...150050normal_gammaNaNNaN250.00013433.013270777.257509
2[34.51079683893437]0.0002330.0000850.0000550.025118NaN0.0[3.288441345698814, 22.89133790738252, 18.6152...kl_pp0.0...150050normal_gammaNaNNaN250.00014037.224562833.980804
\n", + "

3 rows × 31 columns

\n", + "
" + ], + "text/plain": [ + " solution \\\n", + "uuid replication \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 [26.504853109137507] \n", + " 1 [29.879991219010936] \n", + " 2 [34.51079683893437] \n", + "\n", + " dgp_time likelihood_time \\\n", + "uuid replication \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 0.000326 0.000119 \n", + " 1 0.000213 0.000081 \n", + " 2 0.000233 0.000085 \n", + "\n", + " posterior_time solve_time \\\n", + "uuid replication \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 0.000081 0.042336 \n", + " 1 0.000053 0.027507 \n", + " 2 0.000055 0.025118 \n", + "\n", + " setup_time \\\n", + "uuid replication \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 NaN \n", + " 1 NaN \n", + " 2 NaN \n", + "\n", + " log_partition_constant \\\n", + "uuid replication \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 0.0 \n", + " 1 0.0 \n", + " 2 0.0 \n", + "\n", + " out_of_sample_cost \\\n", + "uuid replication \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 [8.370599215733552, 97.27825277087481, 24.4703... \n", + " 1 [14.395708241482502, 22.90806081601392, 64.485... \n", + " 2 [3.288441345698814, 22.89133790738252, 18.6152... \n", + "\n", + " algorithm contamination \\\n", + "uuid replication \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 kl_pp 0.0 \n", + " 1 kl_pp 0.0 \n", + " 2 kl_pp 0.0 \n", + "\n", + " ... num_posterior_samples \\\n", + "uuid replication ... \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 ... 1 \n", + " 1 ... 1 \n", + " 2 ... 1 \n", + "\n", + " num_replications \\\n", + "uuid replication \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 500 \n", + " 1 500 \n", + " 2 500 \n", + "\n", + " num_test_observations \\\n", + "uuid replication \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 50 \n", + " 1 50 \n", + " 2 50 \n", + "\n", + " posterior kde_epsilon \\\n", + "uuid replication \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 normal_gamma NaN \n", + " 1 normal_gamma NaN \n", + " 2 normal_gamma NaN \n", + "\n", + " n_splits num_total_samples \\\n", + "uuid replication \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 NaN 25 \n", + " 1 NaN 25 \n", + " 2 NaN 25 \n", + "\n", + " sample_time in_group_mean \\\n", + "uuid replication \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 0.000200 44.554176 \n", + " 1 0.000134 33.013270 \n", + " 2 0.000140 37.224562 \n", + "\n", + " in_group_var \n", + "uuid replication \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 1543.100317 \n", + " 1 777.257509 \n", + " 2 833.980804 \n", + "\n", + "[3 rows x 31 columns]" + ] + }, + "execution_count": 7, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "results_df = preprocess_results_df(all_results_df, filter_dgp)\n", + "results_df.head(3)" + ] + }, + { + "cell_type": "code", + "execution_count": 8, + "metadata": {}, + "outputs": [], + "source": [ + "# NOTE if you are visualising the 'num_observations' experiment, uncomment below line\n", + "# results_df = results_df.loc[(results_df[\"num_total_samples\"] == 100) | (results_df[\"algorithm\"] == \"kl_empirical\")]" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/dcs/pg20/u1508153/projects/mis-dro-code/mis_dro/results.py:63: FutureWarning: The provided callable is currently using SeriesGroupBy.mean. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"mean\" instead.\n", + " agg_df = gb.agg(\n", + "/dcs/pg20/u1508153/projects/mis-dro-code/mis_dro/results.py:63: FutureWarning: The provided callable is currently using SeriesGroupBy.std. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"std\" instead.\n", + " agg_df = gb.agg(\n", + "/dcs/pg20/u1508153/projects/mis-dro-code/mis_dro/results.py:63: FutureWarning: The provided callable is currently using SeriesGroupBy.mean. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"mean\" instead.\n", + " agg_df = gb.agg(\n", + "/dcs/pg20/u1508153/projects/mis-dro-code/mis_dro/results.py:63: FutureWarning: The provided callable is currently using SeriesGroupBy.std. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"std\" instead.\n", + " agg_df = gb.agg(\n" + ] + }, + { + "data": { + "text/html": [ + "
\n", + "\n", + "\n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + " \n", + "
out_of_sample_meanout_of_sample_varsum_of_in_group_varvar_of_in_group_meanmean_solve_timestd_solve_timemean_sample_timestd_sample_time
algorithmdgpepsiloninferencekde_epsilonnum_total_samplesnum_observations
\n", + "
" + ], + "text/plain": [ + "Empty DataFrame\n", + "Columns: [out_of_sample_mean, out_of_sample_var, sum_of_in_group_var, var_of_in_group_mean, mean_solve_time, std_solve_time, mean_sample_time, std_sample_time]\n", + "Index: []" + ] + }, + "execution_count": 13, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "agg_df = get_agg_df(results_df, [\"algorithm\", \"dgp\", \"epsilon\", \"inference\", \"num_total_samples\", \"num_observations\"])\n", + "agg_df" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Plot mean-variance trade-off\n", + "\n", + "For each DGP, we plot the out-of-sample mean $\\hat{\\mu}_M(\\epsilon)$ and variance $\\hat{\\sigma}_M(\\epsilon)$ of Bayesian DRO with different posteriors." + ] + }, + { + "cell_type": "code", + "execution_count": 12, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "All BDRO points are Pareto dominated for M = 25 ? True\n", + "All BDRO points are Pareto dominated for M = 100 ? True\n", + "All BDRO points are Pareto dominated for M = 900 ? False\n" + ] + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "total_samples_list = agg_df.loc[~(agg_df.index.get_level_values(\"algorithm\").isin((\"kl_empirical\", \"wasserstein_empirical\")))].index.get_level_values(\"num_total_samples\").unique().tolist()\n", + "# num_observations_list = agg_df.index.get_level_values(\"num_observations\").unique().tolist()\n", + "dgp_list = agg_df.index.get_level_values(\"dgp\").unique().tolist()\n", + "\n", + "nrows = len(dgp_list)\n", + "ncols = len(total_samples_list)\n", + "# ncols = len(num_observations_list)\n", + "trim_epsilon = 1.0\n", + "\n", + "fig, axes = plt.subplots(nrows=nrows, ncols=ncols, sharex=True, sharey=True, figsize=(5*ncols, nrows*4))\n", + "fig.subplots_adjust(wspace=0.05)\n", + "pd.options.mode.chained_assignment = None # default='warn'\n", + "for i, dgp in enumerate(dgp_list):\n", + " for j, total_samples in enumerate(total_samples_list):\n", + " # for j, num_observations in enumerate(num_observations_list):\n", + "\n", + " # NOTE empirical KL doesn't sample, so we plot it before filtering by total_samples\n", + " # mean_variance_plot(axes[j], agg_df.loc[\"kl_empirical\", dgp, :trim_epsilon, \"empirical\", :, :], **algorithm_inference_style(\"kl_empirical\", \"empirical\", label_inference=False), special_epsilons=[0.3], offset=0.1)\n", + " # mean_variance_plot(axes[j], agg_df.loc[\"kl_empirical\", dgp, :trim_epsilon, \"empirical\", num_observations, :], **algorithm_inference_style(\"kl_empirical\", \"empirical\", label_inference=False), special_epsilons=[0.3], offset=0.1)\n", + " # mean_variance_plot(axes[j], agg_df.loc[\"wasserstein_empirical\", dgp, :, \"empirical\", :, :], **algorithm_inference_style(\"wasserstein_empirical\", \"empirical\", label_inference=False), special_epsilons=[0.3], offset=0.1)\n", + "\n", + "\n", + " # filter by the total samples and DGP\n", + " axis_df = agg_df.loc[:, dgp, :, :, total_samples, :]\n", + " # axis_df = agg_df.loc[:, dgp, :, :, :, num_observations]\n", + " axis_df.loc[:, \"is_pareto_front\"] = is_minimise_pareto_front(axis_df[\"out_of_sample_var\"].values, axis_df[\"out_of_sample_mean\"].values)\n", + " for algorithm, inference in set(zip(axis_df.index.get_level_values(\"algorithm\"), axis_df.index.get_level_values(\"inference\"))):\n", + " df = axis_df.loc[algorithm, :trim_epsilon, :, :]\n", + "\n", + " if algorithm == \"kl_bdro\":\n", + " print(\"All BDRO points are Pareto dominated for M =\", total_samples, \"?\", not df[\"is_pareto_front\"].any())\n", + "\n", + "\n", + " for k in range(j, len(total_samples_list)):\n", + " # for k in range(j, len(num_observations_list)):\n", + " alpha = 1.0\n", + " is_labelled=True\n", + " if j != k:\n", + " alpha = 0.2\n", + " is_labelled = False\n", + " mean_variance_plot(axes[k], df, **algorithm_inference_style(algorithm, inference, label_inference=False), offset=0.0, is_labelled=is_labelled, alpha=alpha)\n", + " if j == 0:\n", + " handles, labels = axes[j].get_legend_handles_labels()\n", + " # algorithm_order = [AlgorithmName.kl_dro_bas.value, AlgorithmName.kl_pp.value, AlgorithmName.kl_bdro.value, AlgorithmName.kl_empirical.value, AlgorithmName.wasserstein_empirical.value]\n", + " algorithm_order = [AlgorithmName.kl_dro_bas.value, AlgorithmName.kl_pp.value, AlgorithmName.kl_bdro.value]\n", + " order = [list(labels).index(a) for a in algorithm_order]\n", + " axes[j].legend([handles[idx] for idx in order],[labels[idx] for idx in order])\n", + " axes[j].set_ylabel(\"Out-of-sample mean, $m(\\epsilon)$\")\n", + " axes[j].set_title(\"$M$\" + f\"={total_samples} with {NiceNameDGP[filter_dgp]}\")\n", + " # axes[j].set_title(\"$n$\" + f\"={num_observations} with {NiceNameDGP[filter_dgp]}\")\n", + " axes[j].set_xlabel(\"Out-of-sample variance, $v(\\epsilon)$\")\n", + "\n", + "# fig.savefig(f\"/Users/patrick/Experiments/misdro/2025_01_21_paper_bas_experiments/{experiment_name}_{filter_dgp}_empirical.pdf\", bbox_inches=\"tight\")" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## Solve & sampling time" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "\\begin{tabular}{llllllll}\n", + "\\toprule\n", + " & & \\multicolumn{3}{r}{str_solve_time} & \\multicolumn{3}{r}{str_sample_time} \\\\\n", + " & & kl_dro_bas & kl_pp & kl_bdro & kl_dro_bas & kl_pp & kl_bdro \\\\\n", + "dgp & num_total_samples & & & & & & \\\\\n", + "\\midrule\n", + "\\multirow[t]{4}{*}{exponential} & 20 & NaN & NaN & NaN & NaN & NaN & NaN \\\\\n", + " & 25 & 0.024 (0.003) & 0.024 (0.003) & 0.068 (0.012) & 0.097 (0.007) & 0.117 (0.009) & 0.360 (0.018) \\\\\n", + " & 100 & 0.036 (0.003) & 0.037 (0.003) & 0.148 (0.020) & 0.099 (0.007) & 0.122 (0.015) & 0.614 (0.045) \\\\\n", + " & 900 & 0.422 (0.030) & 0.428 (0.032) & 0.724 (0.040) & 0.114 (0.010) & 0.155 (0.013) & 1.611 (0.117) \\\\\n", + "\\cline{1-8}\n", + "\\bottomrule\n", + "\\end{tabular}\n", + "\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/dcs/pg20/u1508153/projects/mis-dro-code/mis_dro/plot.py:219: FutureWarning: The provided callable is currently using SeriesGroupBy.mean. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"mean\" instead.\n", + " agg_df = gb.agg(\n", + "/dcs/pg20/u1508153/projects/mis-dro-code/mis_dro/plot.py:219: FutureWarning: The provided callable is currently using SeriesGroupBy.std. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"std\" instead.\n", + " agg_df = gb.agg(\n" + ] + } + ], + "source": [ + "solve_time_df = get_agg_df(results_df, [\"algorithm\", \"dgp\", \"num_total_samples\"])[[\"mean_solve_time\", \"std_solve_time\", \"mean_sample_time\", \"std_sample_time\"]]\n", + "\n", + "decimal_places = 3\n", + "format_string = \"{:.3f}\"\n", + "solve_time_df[\"str_solve_time\"] = solve_time_df[[\"mean_solve_time\", \"std_solve_time\"]].apply(\n", + " lambda x: format_string.format(np.round(x[\"mean_solve_time\"], decimal_places)) + \" (\" + format_string.format(np.round(x[\"std_solve_time\"], decimal_places)) + \")\", axis=1\n", + ")\n", + "solve_time_df[\"str_sample_time\"] = solve_time_df[[\"mean_sample_time\", \"std_sample_time\"]].apply(\n", + " lambda x: format_string.format(np.round(1000 * x[\"mean_sample_time\"], decimal_places)) + \" (\" + format_string.format(np.round(1000 * x[\"std_sample_time\"], decimal_places)) + \")\", axis=1\n", + ")\n", + "# solve_time_df.reindex(pd.MultiIndex.from_product(np.repeat([\"kl_dro_bas\", \"kl_pp\", \"kl_bdro\", \"kl_empirical\"], 3), solve_time_df.index.get_level_values(\"num_total_samples\"), names=['algorithm', 'num_total_samples']))\n", + "# solve_time_df = solve_time_df.reindex([\"kl_dro_bas\", \"kl_pp\", \"kl_bdro\", \"kl_empirical\"])\n", + "# solve_time_df.index = solve_time_df.index.map(lambda x: AlgorithmName[x].value)\n", + "print(solve_time_df[[\"str_solve_time\", \"str_sample_time\"]].unstack(level=\"algorithm\").reindex(columns=pd.MultiIndex.from_product([[\"str_solve_time\", \"str_sample_time\"], [\"kl_dro_bas\", \"kl_pp\", \"kl_bdro\"]])).to_latex())" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# Plot sum of in-group variance and variance of in-group mean\n", + "Let $m$ be the number of test observations.\n", + "Let $k$ be the number of replications.\n", + "Let $\\xi_{ij}$ be the $i^{\\text{th}}$ test observation for replication $j$.\n", + "For a given replication $j$, the in-group mean and variance is\n", + "$$\\mu_j = \\frac{1}{m}\\sum_{i=1}^m f(x_j, \\xi_{ij}), \\hspace{2em} v_j = \\frac{1}{m-1}\\sum_{i=1}^m \\left( f(x_j, \\xi_{ij}) - \\mu_j \\right)^2$$\n", + "\n", + "We define the mean across all replications and all test observations as\n", + "$$\\bar{\\mu} = \\frac{1}{mk} \\sum_{j=1}^k \\sum_{i=1}^m f(x_j, \\xi_{ij}).$$\n", + "\n", + "The total variance is defined by $$\\frac{1}{mk-1}\\sum_{j=1}^k \\sum_{i=1}^m \\left( f(x_j, \\xi_{ij}) - \\bar{\\mu} \\right)^2.$$\n", + "This can be decomposed into two terms.\n", + "The first is a constant times the sum of the in-group variances:\n", + "$$\\frac{m-1}{km - 1} \\sum^k_{j=1} v_j,$$\n", + "and the second term is a constant times the variance of the in-group means:\n", + "$$\\frac{m(k-1)}{km - 1} \\text{Var}(\\mu_j).$$\n" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [ + { + "ename": "KeyError", + "evalue": "25", + "output_type": "error", + "traceback": [ + "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", + "\u001b[0;31mKeyError\u001b[0m Traceback (most recent call last)", + "File \u001b[0;32m~/.conda/envs/mis-dro/lib/python3.11/site-packages/pandas/core/indexes/base.py:3791\u001b[0m, in \u001b[0;36mIndex.get_loc\u001b[0;34m(self, key)\u001b[0m\n\u001b[1;32m 3790\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[0;32m-> 3791\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_engine\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget_loc\u001b[49m\u001b[43m(\u001b[49m\u001b[43mcasted_key\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 3792\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mKeyError\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m err:\n", + "File \u001b[0;32mindex.pyx:152\u001b[0m, in \u001b[0;36mpandas._libs.index.IndexEngine.get_loc\u001b[0;34m()\u001b[0m\n", + "File \u001b[0;32mindex.pyx:181\u001b[0m, in \u001b[0;36mpandas._libs.index.IndexEngine.get_loc\u001b[0;34m()\u001b[0m\n", + "File \u001b[0;32mpandas/_libs/hashtable_class_helper.pxi:7080\u001b[0m, in \u001b[0;36mpandas._libs.hashtable.PyObjectHashTable.get_item\u001b[0;34m()\u001b[0m\n", + "File \u001b[0;32mpandas/_libs/hashtable_class_helper.pxi:7088\u001b[0m, in \u001b[0;36mpandas._libs.hashtable.PyObjectHashTable.get_item\u001b[0;34m()\u001b[0m\n", + "\u001b[0;31mKeyError\u001b[0m: 25", + "\nThe above exception was the direct cause of the following exception:\n", + "\u001b[0;31mKeyError\u001b[0m Traceback (most recent call last)", + "Cell \u001b[0;32mIn[8], line 21\u001b[0m\n\u001b[1;32m 19\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m i, var_col \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28menumerate\u001b[39m([\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mout_of_sample_var\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124msum_of_in_group_var\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mvar_of_in_group_mean\u001b[39m\u001b[38;5;124m\"\u001b[39m]):\n\u001b[1;32m 20\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m j, total_samples \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28menumerate\u001b[39m(total_samples_list):\n\u001b[0;32m---> 21\u001b[0m axis_df \u001b[38;5;241m=\u001b[39m \u001b[43magg_df\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mloc\u001b[49m\u001b[43m[\u001b[49m\u001b[43m:\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mdgp\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43m:\u001b[49m\u001b[43mtrim_epsilon\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mtotal_samples\u001b[49m\u001b[43m]\u001b[49m\n\u001b[1;32m 22\u001b[0m axis_df\u001b[38;5;241m.\u001b[39mloc[:, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mis_pareto_front\u001b[39m\u001b[38;5;124m\"\u001b[39m] \u001b[38;5;241m=\u001b[39m is_minimise_pareto_front(axis_df[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mout_of_sample_var\u001b[39m\u001b[38;5;124m\"\u001b[39m]\u001b[38;5;241m.\u001b[39mvalues, axis_df[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mout_of_sample_mean\u001b[39m\u001b[38;5;124m\"\u001b[39m]\u001b[38;5;241m.\u001b[39mvalues)\n\u001b[1;32m 23\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m algorithm \u001b[38;5;129;01min\u001b[39;00m axis_df\u001b[38;5;241m.\u001b[39mindex\u001b[38;5;241m.\u001b[39mget_level_values(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124malgorithm\u001b[39m\u001b[38;5;124m\"\u001b[39m)\u001b[38;5;241m.\u001b[39munique():\n", + "File \u001b[0;32m~/.conda/envs/mis-dro/lib/python3.11/site-packages/pandas/core/indexing.py:1147\u001b[0m, in \u001b[0;36m_LocationIndexer.__getitem__\u001b[0;34m(self, key)\u001b[0m\n\u001b[1;32m 1145\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_is_scalar_access(key):\n\u001b[1;32m 1146\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mobj\u001b[38;5;241m.\u001b[39m_get_value(\u001b[38;5;241m*\u001b[39mkey, takeable\u001b[38;5;241m=\u001b[39m\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_takeable)\n\u001b[0;32m-> 1147\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_getitem_tuple\u001b[49m\u001b[43m(\u001b[49m\u001b[43mkey\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 1148\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m 1149\u001b[0m \u001b[38;5;66;03m# we by definition only have the 0th axis\u001b[39;00m\n\u001b[1;32m 1150\u001b[0m axis \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39maxis \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;241m0\u001b[39m\n", + "File \u001b[0;32m~/.conda/envs/mis-dro/lib/python3.11/site-packages/pandas/core/indexing.py:1330\u001b[0m, in \u001b[0;36m_LocIndexer._getitem_tuple\u001b[0;34m(self, tup)\u001b[0m\n\u001b[1;32m 1328\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m suppress(IndexingError):\n\u001b[1;32m 1329\u001b[0m tup \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_expand_ellipsis(tup)\n\u001b[0;32m-> 1330\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_getitem_lowerdim\u001b[49m\u001b[43m(\u001b[49m\u001b[43mtup\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 1332\u001b[0m \u001b[38;5;66;03m# no multi-index, so validate all of the indexers\u001b[39;00m\n\u001b[1;32m 1333\u001b[0m tup \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_validate_tuple_indexer(tup)\n", + "File \u001b[0;32m~/.conda/envs/mis-dro/lib/python3.11/site-packages/pandas/core/indexing.py:1015\u001b[0m, in \u001b[0;36m_LocationIndexer._getitem_lowerdim\u001b[0;34m(self, tup)\u001b[0m\n\u001b[1;32m 1013\u001b[0m \u001b[38;5;66;03m# we may have a nested tuples indexer here\u001b[39;00m\n\u001b[1;32m 1014\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_is_nested_tuple_indexer(tup):\n\u001b[0;32m-> 1015\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_getitem_nested_tuple\u001b[49m\u001b[43m(\u001b[49m\u001b[43mtup\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 1017\u001b[0m \u001b[38;5;66;03m# we maybe be using a tuple to represent multiple dimensions here\u001b[39;00m\n\u001b[1;32m 1018\u001b[0m ax0 \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mobj\u001b[38;5;241m.\u001b[39m_get_axis(\u001b[38;5;241m0\u001b[39m)\n", + "File \u001b[0;32m~/.conda/envs/mis-dro/lib/python3.11/site-packages/pandas/core/indexing.py:1114\u001b[0m, in \u001b[0;36m_LocationIndexer._getitem_nested_tuple\u001b[0;34m(self, tup)\u001b[0m\n\u001b[1;32m 1111\u001b[0m \u001b[38;5;66;03m# this is a series with a multi-index specified a tuple of\u001b[39;00m\n\u001b[1;32m 1112\u001b[0m \u001b[38;5;66;03m# selectors\u001b[39;00m\n\u001b[1;32m 1113\u001b[0m axis \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39maxis \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;241m0\u001b[39m\n\u001b[0;32m-> 1114\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_getitem_axis\u001b[49m\u001b[43m(\u001b[49m\u001b[43mtup\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43maxis\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43maxis\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 1116\u001b[0m \u001b[38;5;66;03m# handle the multi-axis by taking sections and reducing\u001b[39;00m\n\u001b[1;32m 1117\u001b[0m \u001b[38;5;66;03m# this is iterative\u001b[39;00m\n\u001b[1;32m 1118\u001b[0m obj \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mobj\n", + "File \u001b[0;32m~/.conda/envs/mis-dro/lib/python3.11/site-packages/pandas/core/indexing.py:1386\u001b[0m, in \u001b[0;36m_LocIndexer._getitem_axis\u001b[0;34m(self, key, axis)\u001b[0m\n\u001b[1;32m 1384\u001b[0m \u001b[38;5;66;03m# nested tuple slicing\u001b[39;00m\n\u001b[1;32m 1385\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m is_nested_tuple(key, labels):\n\u001b[0;32m-> 1386\u001b[0m locs \u001b[38;5;241m=\u001b[39m \u001b[43mlabels\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget_locs\u001b[49m\u001b[43m(\u001b[49m\u001b[43mkey\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 1387\u001b[0m indexer \u001b[38;5;241m=\u001b[39m [\u001b[38;5;28mslice\u001b[39m(\u001b[38;5;28;01mNone\u001b[39;00m)] \u001b[38;5;241m*\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mndim\n\u001b[1;32m 1388\u001b[0m indexer[axis] \u001b[38;5;241m=\u001b[39m locs\n", + "File \u001b[0;32m~/.conda/envs/mis-dro/lib/python3.11/site-packages/pandas/core/indexes/multi.py:3419\u001b[0m, in \u001b[0;36mMultiIndex.get_locs\u001b[0;34m(self, seq)\u001b[0m\n\u001b[1;32m 3415\u001b[0m \u001b[38;5;28;01mcontinue\u001b[39;00m\n\u001b[1;32m 3417\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m 3418\u001b[0m \u001b[38;5;66;03m# a slice or a single label\u001b[39;00m\n\u001b[0;32m-> 3419\u001b[0m lvl_indexer \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_get_level_indexer\u001b[49m\u001b[43m(\u001b[49m\u001b[43mk\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mlevel\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mi\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mindexer\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mindexer\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 3421\u001b[0m \u001b[38;5;66;03m# update indexer\u001b[39;00m\n\u001b[1;32m 3422\u001b[0m lvl_indexer \u001b[38;5;241m=\u001b[39m _to_bool_indexer(lvl_indexer)\n", + "File \u001b[0;32m~/.conda/envs/mis-dro/lib/python3.11/site-packages/pandas/core/indexes/multi.py:3276\u001b[0m, in \u001b[0;36mMultiIndex._get_level_indexer\u001b[0;34m(self, key, level, indexer)\u001b[0m\n\u001b[1;32m 3273\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mslice\u001b[39m(i, j, step)\n\u001b[1;32m 3275\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m-> 3276\u001b[0m idx \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_get_loc_single_level_index\u001b[49m\u001b[43m(\u001b[49m\u001b[43mlevel_index\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mkey\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 3278\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m level \u001b[38;5;241m>\u001b[39m \u001b[38;5;241m0\u001b[39m \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_lexsort_depth \u001b[38;5;241m==\u001b[39m \u001b[38;5;241m0\u001b[39m:\n\u001b[1;32m 3279\u001b[0m \u001b[38;5;66;03m# Desired level is not sorted\u001b[39;00m\n\u001b[1;32m 3280\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28misinstance\u001b[39m(idx, \u001b[38;5;28mslice\u001b[39m):\n\u001b[1;32m 3281\u001b[0m \u001b[38;5;66;03m# test_get_loc_partial_timestamp_multiindex\u001b[39;00m\n", + "File \u001b[0;32m~/.conda/envs/mis-dro/lib/python3.11/site-packages/pandas/core/indexes/multi.py:2865\u001b[0m, in \u001b[0;36mMultiIndex._get_loc_single_level_index\u001b[0;34m(self, level_index, key)\u001b[0m\n\u001b[1;32m 2863\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m1\u001b[39m\n\u001b[1;32m 2864\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m-> 2865\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mlevel_index\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget_loc\u001b[49m\u001b[43m(\u001b[49m\u001b[43mkey\u001b[49m\u001b[43m)\u001b[49m\n", + "File \u001b[0;32m~/.conda/envs/mis-dro/lib/python3.11/site-packages/pandas/core/indexes/base.py:3798\u001b[0m, in \u001b[0;36mIndex.get_loc\u001b[0;34m(self, key)\u001b[0m\n\u001b[1;32m 3793\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28misinstance\u001b[39m(casted_key, \u001b[38;5;28mslice\u001b[39m) \u001b[38;5;129;01mor\u001b[39;00m (\n\u001b[1;32m 3794\u001b[0m \u001b[38;5;28misinstance\u001b[39m(casted_key, abc\u001b[38;5;241m.\u001b[39mIterable)\n\u001b[1;32m 3795\u001b[0m \u001b[38;5;129;01mand\u001b[39;00m \u001b[38;5;28many\u001b[39m(\u001b[38;5;28misinstance\u001b[39m(x, \u001b[38;5;28mslice\u001b[39m) \u001b[38;5;28;01mfor\u001b[39;00m x \u001b[38;5;129;01min\u001b[39;00m casted_key)\n\u001b[1;32m 3796\u001b[0m ):\n\u001b[1;32m 3797\u001b[0m \u001b[38;5;28;01mraise\u001b[39;00m InvalidIndexError(key)\n\u001b[0;32m-> 3798\u001b[0m \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mKeyError\u001b[39;00m(key) \u001b[38;5;28;01mfrom\u001b[39;00m \u001b[38;5;21;01merr\u001b[39;00m\n\u001b[1;32m 3799\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mTypeError\u001b[39;00m:\n\u001b[1;32m 3800\u001b[0m \u001b[38;5;66;03m# If we have a listlike key, _check_indexing_error will raise\u001b[39;00m\n\u001b[1;32m 3801\u001b[0m \u001b[38;5;66;03m# InvalidIndexError. Otherwise we fall through and re-raise\u001b[39;00m\n\u001b[1;32m 3802\u001b[0m \u001b[38;5;66;03m# the TypeError.\u001b[39;00m\n\u001b[1;32m 3803\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_check_indexing_error(key)\n", + "\u001b[0;31mKeyError\u001b[0m: 25" + ] + }, + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "total_samples_list = agg_df.index.get_level_values(\"num_total_samples\").unique().tolist()\n", + "dgp_list = agg_df.index.get_level_values(\"dgp\").unique().tolist()\n", + "\n", + "nrows = 3\n", + "ncols = len(total_samples_list)\n", + "trim_epsilon = 1.0\n", + "\n", + "fig, axes = plt.subplots(nrows=nrows, ncols=ncols, sharex=True, sharey=True, figsize=(5*ncols, nrows*5))\n", + "fig.subplots_adjust(wspace=0.05)\n", + "pd.options.mode.chained_assignment = None # default='warn'\n", + "fig.suptitle(f\"{NiceNameDGP[filter_dgp]} newsvendor\", fontsize=16)\n", + "\n", + "nice_var_names = {\n", + " \"out_of_sample_var\": \"Total variance\",\n", + " \"sum_of_in_group_var\": r\"$\\frac{m-1}{km - 1} \\sum^k_{j=1} v_j$\",\n", + " \"var_of_in_group_mean\": r\"$\\frac{m(k-1)}{km - 1} \\text{Var}(\\mu_j)$\",\n", + "}\n", + "\n", + "for i, var_col in enumerate([\"out_of_sample_var\", \"sum_of_in_group_var\", \"var_of_in_group_mean\"]):\n", + " for j, total_samples in enumerate(total_samples_list):\n", + " axis_df = agg_df.loc[:, dgp, :trim_epsilon, total_samples]\n", + " axis_df.loc[:, \"is_pareto_front\"] = is_minimise_pareto_front(axis_df[\"out_of_sample_var\"].values, axis_df[\"out_of_sample_mean\"].values)\n", + " for algorithm in axis_df.index.get_level_values(\"algorithm\").unique():\n", + " df = axis_df.loc[algorithm, :]\n", + " alpha = 1.0\n", + " is_labelled=True\n", + " mean_variance_plot(axes[i][j], df, **algorithm_inference_style(algorithm, \"bayes\", label_inference=False), offset=0.0, is_labelled=is_labelled, alpha=alpha, var_col=var_col)\n", + " if j == 0:\n", + " axes[i][j].legend()\n", + " axes[i][j].set_ylabel(\"Out-of-sample mean\")\n", + " axes[i][j].set_title(nice_var_names[var_col] + \" for $M$\" + f\"={total_samples}\")\n", + " axes[i][j].set_xlabel(\"Out-of-sample variance\")\n", + "\n", + "# fig.savefig(f\"/Users/patrick/Experiments/misdro/paper_bas_figures/{experiment_name}_{filter_dgp}.pdf\", bbox_inches=\"tight\")" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# Look at the solution" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "df = results_df.copy()\n", + "df[\"solution\"] = df[\"solution\"].map(lambda x: x[4])\n", + "nrows = len(dgp_list)\n", + "ncols = len(total_samples_list)\n", + "fig, axes = plt.subplots(ncols=ncols, nrows=nrows, sharex=True, sharey=True, figsize=(5*ncols, nrows*5))\n", + "solution_df = df.groupby([\"algorithm\", \"dgp\", \"epsilon\", \"num_total_samples\"]).agg({\"solution\": [\"mean\", \"std\"]})\n", + "for i, dgp in enumerate(dgp_list):\n", + " for j, total_samples in enumerate(total_samples_list):\n", + " axis_df = solution_df.loc[:, dgp, :, total_samples]\n", + " for algorithm in axis_df.index.get_level_values(\"algorithm\").unique():\n", + " axes[j].errorbar(axis_df.loc[algorithm, :].index.get_level_values(\"epsilon\"), axis_df.loc[algorithm, :][\"solution\"][\"mean\"], yerr=axis_df.loc[algorithm, :][\"solution\"][\"std\"], **algorithm_inference_style(algorithm, \"bayes\"))\n", + " axes[j].set_xscale(\"log\")\n", + " axes[j].set_title(\"$M$\" + f\"={total_samples} with {NiceNameDGP[filter_dgp]}\")\n", + " axes[j].set_xlabel(\"Epsilon\")\n", + " if j == 0:\n", + " axes[j].set_ylabel(\"Solution\")" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [] + } + ], + "metadata": { + "kernelspec": { + "display_name": "mis-dro", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.11.7" + }, + "orig_nbformat": 4 + }, + "nbformat": 4, + "nbformat_minor": 2 +} diff --git a/notebooks/newsvendor_experiment.ipynb b/notebooks/newsvendor_experiment.ipynb index 0e5b86a..a929236 100644 --- a/notebooks/newsvendor_experiment.ipynb +++ b/notebooks/newsvendor_experiment.ipynb @@ -2,7 +2,7 @@ "cells": [ { "cell_type": "code", - "execution_count": 1, + "execution_count": 2, "metadata": {}, "outputs": [], "source": [ @@ -16,6 +16,8 @@ "from mis_dro.plot import *\n", "from mis_dro.experiments import ExperimentName\n", "\n", + "from mis_dro.results import preprocess_results_df, is_minimise_pareto_front, get_agg_df\n", + "\n", "from matplotlib import rc\n", "rc('font', **{'family': 'serif', 'serif': ['Computer Modern'], \"size\":14})\n", "rc('text', usetex=True)" @@ -23,7 +25,24 @@ }, { "cell_type": "code", - "execution_count": 2, + "execution_count": 3, + "metadata": {}, + "outputs": [], + "source": [ + "experiment_name = ExperimentName.kl_newsvendor_1d\n", + "experiment_dir = Path(f\"/Users/patrick/experiments/misdro/2025_01_21_paper_bas_experiments/{experiment_name.value}\")\n", + "# experiment_dir = Path(f\"/dcs/large/u1508153/misdro/2025_01_21_paper_bas_experiments/{experiment_name.value}\")\n", + "all_results_df = pd.read_csv(experiment_dir / \"results.csv\", index_col=[\"uuid\", \"replication\"])\n", + "\n", + "wass_dir = Path(\"/Users/patrick/Experiments/misdro/2025_01_21_paper_bas_experiments/wasserstein_empirical\")\n", + "wass_df = pd.read_csv(wass_dir / \"results.csv\", index_col=[\"uuid\", \"replication\"])\n", + "\n", + "all_results_df = pd.concat([all_results_df, wass_df])" + ] + }, + { + "cell_type": "code", + "execution_count": 11, "metadata": {}, "outputs": [ { @@ -59,16 +78,16 @@ " algorithm\n", " contamination\n", " ...\n", - " inference\n", - " lengthscale\n", - " likelihood\n", - " njobs\n", - " num_likelihood_samples\n", " num_observations\n", " num_posterior_samples\n", " num_replications\n", " num_test_observations\n", " posterior\n", + " use_cv_epsilon\n", + " num_total_samples\n", + " sample_time\n", + " in_group_mean\n", + " in_group_var\n", " \n", " \n", " uuid\n", @@ -98,7 +117,7 @@ " \n", " \n", " \n", - " 10aee0fd-c2c4-448d-995a-b307828a920e\n", + " 10aee0fd-c2c4-448d-995a-b307828a920e\n", " 0\n", " [26.504853109137507]\n", " 0.000326\n", @@ -111,16 +130,16 @@ " kl_pp\n", " 0.0\n", " ...\n", - " bayes\n", - " -1.0\n", - " normal\n", - " 1\n", - " 25\n", " 20\n", " 1\n", " 500\n", " 50\n", " normal_gamma\n", + " False\n", + " 25\n", + " 0.000200\n", + " 44.554176\n", + " 1543.100317\n", " \n", " \n", " 1\n", @@ -135,16 +154,16 @@ " kl_pp\n", " 0.0\n", " ...\n", - " bayes\n", - " -1.0\n", - " normal\n", - " 1\n", - " 25\n", " 20\n", " 1\n", " 500\n", " 50\n", " normal_gamma\n", + " False\n", + " 25\n", + " 0.000134\n", + " 33.013270\n", + " 777.257509\n", " \n", " \n", " 2\n", @@ -159,214 +178,20 @@ " kl_pp\n", " 0.0\n", " ...\n", - " bayes\n", - " -1.0\n", - " normal\n", - " 1\n", - " 25\n", " 20\n", " 1\n", " 500\n", " 50\n", " normal_gamma\n", - " \n", - " \n", - " 3\n", - " [41.21168988315753]\n", - " 0.000217\n", - " 0.000082\n", - " 0.000050\n", - " 0.026851\n", - " NaN\n", - " 0.0\n", - " [47.90728269717266, 2.2604441130882265, 32.281...\n", - " kl_pp\n", - " 0.0\n", - " ...\n", - " bayes\n", - " -1.0\n", - " normal\n", - " 1\n", + " False\n", " 25\n", - " 20\n", - " 1\n", - " 500\n", - " 50\n", - " normal_gamma\n", - " \n", - " \n", - " 4\n", - " [39.2247838013642]\n", - " 0.000201\n", - " 0.000079\n", - " 0.000053\n", - " 0.025586\n", - " NaN\n", - " 0.0\n", - " [62.36290173196352, 62.83479182385227, 31.2686...\n", - " kl_pp\n", - " 0.0\n", - " ...\n", - " bayes\n", - " -1.0\n", - " normal\n", - " 1\n", - " 25\n", - " 20\n", - " 1\n", - " 500\n", - " 50\n", - " normal_gamma\n", - " \n", - " \n", - " ...\n", - " ...\n", - " ...\n", - " ...\n", - " ...\n", - " ...\n", - " ...\n", - " ...\n", - " ...\n", - " ...\n", - " ...\n", - " ...\n", - " ...\n", - " ...\n", - " ...\n", - " ...\n", - " ...\n", - " ...\n", - " ...\n", - " ...\n", - " ...\n", - " ...\n", - " ...\n", - " \n", - " \n", - " ce796100-ef49-466b-b9a2-71d82aff8197\n", - " 495\n", - " [73.30164681370464]\n", - " 0.000378\n", - " 0.000001\n", - " 0.000011\n", - " 0.020793\n", - " NaN\n", - " 0.0\n", - " [208.0543562579382, 123.33136483660306, 166.76...\n", - " kl_empirical\n", - " 0.2\n", - " ...\n", - " empirical\n", - " -1.0\n", - " empirical\n", - " 1\n", - " 20\n", - " 20\n", - " 1\n", - " 500\n", - " 50\n", - " empirical\n", - " \n", - " \n", - " 496\n", - " [73.3680181030691]\n", - " 0.000395\n", - " 0.000001\n", - " 0.000011\n", - " 0.020304\n", - " NaN\n", - " 0.0\n", - " [136.79531028324774, 96.39454384514991, 68.871...\n", - " kl_empirical\n", - " 0.2\n", - " ...\n", - " empirical\n", - " -1.0\n", - " empirical\n", - " 1\n", - " 20\n", - " 20\n", - " 1\n", - " 500\n", - " 50\n", - " empirical\n", - " \n", - " \n", - " 497\n", - " [73.3282067769722]\n", - " 0.000375\n", - " 0.000000\n", - " 0.000011\n", - " 0.019942\n", - " NaN\n", - " 0.0\n", - " [211.4558369712576, 129.0677245179817, 216.601...\n", - " kl_empirical\n", - " 0.2\n", - " ...\n", - " empirical\n", - " -1.0\n", - " empirical\n", - " 1\n", - " 20\n", - " 20\n", - " 1\n", - " 500\n", - " 50\n", - " empirical\n", - " \n", - " \n", - " 498\n", - " [72.9673128897069]\n", - " 0.000363\n", - " 0.000000\n", - " 0.000011\n", - " 0.019926\n", - " NaN\n", - " 0.0\n", - " [216.00904620730623, 184.2192781513036, 91.823...\n", - " kl_empirical\n", - " 0.2\n", - " ...\n", - " empirical\n", - " -1.0\n", - " empirical\n", - " 1\n", - " 20\n", - " 20\n", - " 1\n", - " 500\n", - " 50\n", - " empirical\n", - " \n", - " \n", - " 499\n", - " [88.86478164898362]\n", - " 0.000362\n", - " 0.000001\n", - " 0.000009\n", - " 0.020151\n", - " NaN\n", - " 0.0\n", - " [259.108516359433, 223.44167581175873, 156.812...\n", - " kl_empirical\n", - " 0.2\n", - " ...\n", - " empirical\n", - " -1.0\n", - " empirical\n", - " 1\n", - " 20\n", - " 20\n", - " 1\n", - " 500\n", - " 50\n", - " empirical\n", + " 0.000140\n", + " 37.224562\n", + " 833.980804\n", " \n", " \n", "\n", - "

560000 rows × 25 columns

\n", + "

3 rows × 30 columns

\n", "" ], "text/plain": [ @@ -375,495 +200,109 @@ "10aee0fd-c2c4-448d-995a-b307828a920e 0 [26.504853109137507] \n", " 1 [29.879991219010936] \n", " 2 [34.51079683893437] \n", - " 3 [41.21168988315753] \n", - " 4 [39.2247838013642] \n", - "... ... \n", - "ce796100-ef49-466b-b9a2-71d82aff8197 495 [73.30164681370464] \n", - " 496 [73.3680181030691] \n", - " 497 [73.3282067769722] \n", - " 498 [72.9673128897069] \n", - " 499 [88.86478164898362] \n", "\n", " dgp_time likelihood_time \\\n", "uuid replication \n", "10aee0fd-c2c4-448d-995a-b307828a920e 0 0.000326 0.000119 \n", " 1 0.000213 0.000081 \n", " 2 0.000233 0.000085 \n", - " 3 0.000217 0.000082 \n", - " 4 0.000201 0.000079 \n", - "... ... ... \n", - "ce796100-ef49-466b-b9a2-71d82aff8197 495 0.000378 0.000001 \n", - " 496 0.000395 0.000001 \n", - " 497 0.000375 0.000000 \n", - " 498 0.000363 0.000000 \n", - " 499 0.000362 0.000001 \n", "\n", " posterior_time solve_time \\\n", "uuid replication \n", "10aee0fd-c2c4-448d-995a-b307828a920e 0 0.000081 0.042336 \n", " 1 0.000053 0.027507 \n", " 2 0.000055 0.025118 \n", - " 3 0.000050 0.026851 \n", - " 4 0.000053 0.025586 \n", - "... ... ... \n", - "ce796100-ef49-466b-b9a2-71d82aff8197 495 0.000011 0.020793 \n", - " 496 0.000011 0.020304 \n", - " 497 0.000011 0.019942 \n", - " 498 0.000011 0.019926 \n", - " 499 0.000009 0.020151 \n", "\n", " setup_time \\\n", "uuid replication \n", "10aee0fd-c2c4-448d-995a-b307828a920e 0 NaN \n", " 1 NaN \n", " 2 NaN \n", - " 3 NaN \n", - " 4 NaN \n", - "... ... \n", - "ce796100-ef49-466b-b9a2-71d82aff8197 495 NaN \n", - " 496 NaN \n", - " 497 NaN \n", - " 498 NaN \n", - " 499 NaN \n", "\n", " log_partition_constant \\\n", "uuid replication \n", "10aee0fd-c2c4-448d-995a-b307828a920e 0 0.0 \n", " 1 0.0 \n", " 2 0.0 \n", - " 3 0.0 \n", - " 4 0.0 \n", - "... ... \n", - "ce796100-ef49-466b-b9a2-71d82aff8197 495 0.0 \n", - " 496 0.0 \n", - " 497 0.0 \n", - " 498 0.0 \n", - " 499 0.0 \n", "\n", " out_of_sample_cost \\\n", "uuid replication \n", "10aee0fd-c2c4-448d-995a-b307828a920e 0 [8.370599215733552, 97.27825277087481, 24.4703... \n", " 1 [14.395708241482502, 22.90806081601392, 64.485... \n", " 2 [3.288441345698814, 22.89133790738252, 18.6152... \n", - " 3 [47.90728269717266, 2.2604441130882265, 32.281... \n", - " 4 [62.36290173196352, 62.83479182385227, 31.2686... \n", - "... ... \n", - "ce796100-ef49-466b-b9a2-71d82aff8197 495 [208.0543562579382, 123.33136483660306, 166.76... \n", - " 496 [136.79531028324774, 96.39454384514991, 68.871... \n", - " 497 [211.4558369712576, 129.0677245179817, 216.601... \n", - " 498 [216.00904620730623, 184.2192781513036, 91.823... \n", - " 499 [259.108516359433, 223.44167581175873, 156.812... \n", - "\n", - " algorithm contamination \\\n", - "uuid replication \n", - "10aee0fd-c2c4-448d-995a-b307828a920e 0 kl_pp 0.0 \n", - " 1 kl_pp 0.0 \n", - " 2 kl_pp 0.0 \n", - " 3 kl_pp 0.0 \n", - " 4 kl_pp 0.0 \n", - "... ... ... \n", - "ce796100-ef49-466b-b9a2-71d82aff8197 495 kl_empirical 0.2 \n", - " 496 kl_empirical 0.2 \n", - " 497 kl_empirical 0.2 \n", - " 498 kl_empirical 0.2 \n", - " 499 kl_empirical 0.2 \n", - "\n", - " ... inference lengthscale \\\n", - "uuid replication ... \n", - "10aee0fd-c2c4-448d-995a-b307828a920e 0 ... bayes -1.0 \n", - " 1 ... bayes -1.0 \n", - " 2 ... bayes -1.0 \n", - " 3 ... bayes -1.0 \n", - " 4 ... bayes -1.0 \n", - "... ... ... ... \n", - "ce796100-ef49-466b-b9a2-71d82aff8197 495 ... empirical -1.0 \n", - " 496 ... empirical -1.0 \n", - " 497 ... empirical -1.0 \n", - " 498 ... empirical -1.0 \n", - " 499 ... empirical -1.0 \n", - "\n", - " likelihood njobs \\\n", - "uuid replication \n", - "10aee0fd-c2c4-448d-995a-b307828a920e 0 normal 1 \n", - " 1 normal 1 \n", - " 2 normal 1 \n", - " 3 normal 1 \n", - " 4 normal 1 \n", - "... ... ... \n", - "ce796100-ef49-466b-b9a2-71d82aff8197 495 empirical 1 \n", - " 496 empirical 1 \n", - " 497 empirical 1 \n", - " 498 empirical 1 \n", - " 499 empirical 1 \n", - "\n", - " num_likelihood_samples \\\n", - "uuid replication \n", - "10aee0fd-c2c4-448d-995a-b307828a920e 0 25 \n", - " 1 25 \n", - " 2 25 \n", - " 3 25 \n", - " 4 25 \n", - "... ... \n", - "ce796100-ef49-466b-b9a2-71d82aff8197 495 20 \n", - " 496 20 \n", - " 497 20 \n", - " 498 20 \n", - " 499 20 \n", - "\n", - " num_observations \\\n", - "uuid replication \n", - "10aee0fd-c2c4-448d-995a-b307828a920e 0 20 \n", - " 1 20 \n", - " 2 20 \n", - " 3 20 \n", - " 4 20 \n", - "... ... \n", - "ce796100-ef49-466b-b9a2-71d82aff8197 495 20 \n", - " 496 20 \n", - " 497 20 \n", - " 498 20 \n", - " 499 20 \n", - "\n", - " num_posterior_samples \\\n", - "uuid replication \n", - "10aee0fd-c2c4-448d-995a-b307828a920e 0 1 \n", - " 1 1 \n", - " 2 1 \n", - " 3 1 \n", - " 4 1 \n", - "... ... \n", - "ce796100-ef49-466b-b9a2-71d82aff8197 495 1 \n", - " 496 1 \n", - " 497 1 \n", - " 498 1 \n", - " 499 1 \n", - "\n", - " num_replications \\\n", - "uuid replication \n", - "10aee0fd-c2c4-448d-995a-b307828a920e 0 500 \n", - " 1 500 \n", - " 2 500 \n", - " 3 500 \n", - " 4 500 \n", - "... ... \n", - "ce796100-ef49-466b-b9a2-71d82aff8197 495 500 \n", - " 496 500 \n", - " 497 500 \n", - " 498 500 \n", - " 499 500 \n", - "\n", - " num_test_observations \\\n", - "uuid replication \n", - "10aee0fd-c2c4-448d-995a-b307828a920e 0 50 \n", - " 1 50 \n", - " 2 50 \n", - " 3 50 \n", - " 4 50 \n", - "... ... \n", - "ce796100-ef49-466b-b9a2-71d82aff8197 495 50 \n", - " 496 50 \n", - " 497 50 \n", - " 498 50 \n", - " 499 50 \n", - "\n", - " posterior \n", - "uuid replication \n", - "10aee0fd-c2c4-448d-995a-b307828a920e 0 normal_gamma \n", - " 1 normal_gamma \n", - " 2 normal_gamma \n", - " 3 normal_gamma \n", - " 4 normal_gamma \n", - "... ... \n", - "ce796100-ef49-466b-b9a2-71d82aff8197 495 empirical \n", - " 496 empirical \n", - " 497 empirical \n", - " 498 empirical \n", - " 499 empirical \n", - "\n", - "[560000 rows x 25 columns]" - ] - }, - "execution_count": 2, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "experiment_name = ExperimentName.kl_newsvendor_1d\n", - "experiment_dir = Path(f\"/Users/patrick/experiments/misdro/2025_01_21_paper_bas_experiments/{experiment_name.value}\")\n", - "all_results_df = pd.read_csv(experiment_dir / \"results.csv\", index_col=[\"uuid\", \"replication\"])\n", - "all_results_df\n" - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "metadata": {}, - "outputs": [ - { - "data": { - "text/html": [ - "
\n", - "\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
solutiondgp_timelikelihood_timeposterior_timesolve_timesetup_timelog_partition_constantout_of_sample_costalgorithmcontamination...num_likelihood_samplesnum_observationsnum_posterior_samplesnum_replicationsnum_test_observationsposteriornum_total_samplessample_timein_group_meanin_group_var
uuidreplication
905d2b22-d634-43bd-88ab-301e4c4544ad0[15.806229409672879]0.0007910.0001030.0000560.040409NaN0.0[44.59710870444056, 36.42214922132565, 0.86402...kl_pp0.0...2520150050normal_gamma250.00015931.0933921030.844283
1[16.053047357969728]0.0006690.0000910.0000590.027276NaN0.0[16.08116828746995, 26.181524779389363, 13.106...kl_pp0.0...2520150050normal_gamma250.00015031.794166760.212738
2[15.950415767039745]0.0006150.0000880.0000550.030464NaN0.0[21.64174709524606, 11.129969030564544, 58.907...kl_pp0.0...2520150050normal_gamma250.00014327.914621482.826141
\n", - "

3 rows × 29 columns

\n", - "
" - ], - "text/plain": [ - " solution \\\n", - "uuid replication \n", - "905d2b22-d634-43bd-88ab-301e4c4544ad 0 [15.806229409672879] \n", - " 1 [16.053047357969728] \n", - " 2 [15.950415767039745] \n", - "\n", - " dgp_time likelihood_time \\\n", - "uuid replication \n", - "905d2b22-d634-43bd-88ab-301e4c4544ad 0 0.000791 0.000103 \n", - " 1 0.000669 0.000091 \n", - " 2 0.000615 0.000088 \n", - "\n", - " posterior_time solve_time \\\n", - "uuid replication \n", - "905d2b22-d634-43bd-88ab-301e4c4544ad 0 0.000056 0.040409 \n", - " 1 0.000059 0.027276 \n", - " 2 0.000055 0.030464 \n", - "\n", - " setup_time \\\n", - "uuid replication \n", - "905d2b22-d634-43bd-88ab-301e4c4544ad 0 NaN \n", - " 1 NaN \n", - " 2 NaN \n", - "\n", - " log_partition_constant \\\n", - "uuid replication \n", - "905d2b22-d634-43bd-88ab-301e4c4544ad 0 0.0 \n", - " 1 0.0 \n", - " 2 0.0 \n", - "\n", - " out_of_sample_cost \\\n", - "uuid replication \n", - "905d2b22-d634-43bd-88ab-301e4c4544ad 0 [44.59710870444056, 36.42214922132565, 0.86402... \n", - " 1 [16.08116828746995, 26.181524779389363, 13.106... \n", - " 2 [21.64174709524606, 11.129969030564544, 58.907... \n", "\n", " algorithm contamination \\\n", "uuid replication \n", - "905d2b22-d634-43bd-88ab-301e4c4544ad 0 kl_pp 0.0 \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 kl_pp 0.0 \n", " 1 kl_pp 0.0 \n", " 2 kl_pp 0.0 \n", "\n", - " ... num_likelihood_samples \\\n", - "uuid replication ... \n", - "905d2b22-d634-43bd-88ab-301e4c4544ad 0 ... 25 \n", - " 1 ... 25 \n", - " 2 ... 25 \n", - "\n", - " num_observations \\\n", - "uuid replication \n", - "905d2b22-d634-43bd-88ab-301e4c4544ad 0 20 \n", - " 1 20 \n", - " 2 20 \n", + " ... num_observations \\\n", + "uuid replication ... \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 ... 20 \n", + " 1 ... 20 \n", + " 2 ... 20 \n", "\n", - " num_posterior_samples \\\n", - "uuid replication \n", - "905d2b22-d634-43bd-88ab-301e4c4544ad 0 1 \n", - " 1 1 \n", - " 2 1 \n", + " num_posterior_samples \\\n", + "uuid replication \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 1 \n", + " 1 1 \n", + " 2 1 \n", "\n", " num_replications \\\n", "uuid replication \n", - "905d2b22-d634-43bd-88ab-301e4c4544ad 0 500 \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 500 \n", " 1 500 \n", " 2 500 \n", "\n", " num_test_observations \\\n", "uuid replication \n", - "905d2b22-d634-43bd-88ab-301e4c4544ad 0 50 \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 50 \n", " 1 50 \n", " 2 50 \n", "\n", - " posterior \\\n", - "uuid replication \n", - "905d2b22-d634-43bd-88ab-301e4c4544ad 0 normal_gamma \n", - " 1 normal_gamma \n", - " 2 normal_gamma \n", + " posterior use_cv_epsilon \\\n", + "uuid replication \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 normal_gamma False \n", + " 1 normal_gamma False \n", + " 2 normal_gamma False \n", "\n", " num_total_samples \\\n", "uuid replication \n", - "905d2b22-d634-43bd-88ab-301e4c4544ad 0 25 \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 25 \n", " 1 25 \n", " 2 25 \n", "\n", " sample_time in_group_mean \\\n", "uuid replication \n", - "905d2b22-d634-43bd-88ab-301e4c4544ad 0 0.000159 31.093392 \n", - " 1 0.000150 31.794166 \n", - " 2 0.000143 27.914621 \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 0.000200 44.554176 \n", + " 1 0.000134 33.013270 \n", + " 2 0.000140 37.224562 \n", "\n", " in_group_var \n", "uuid replication \n", - "905d2b22-d634-43bd-88ab-301e4c4544ad 0 1030.844283 \n", - " 1 760.212738 \n", - " 2 482.826141 \n", + "10aee0fd-c2c4-448d-995a-b307828a920e 0 1543.100317 \n", + " 1 777.257509 \n", + " 2 833.980804 \n", "\n", - "[3 rows x 29 columns]" + "[3 rows x 30 columns]" ] }, - "execution_count": 3, + "execution_count": 11, "metadata": {}, "output_type": "execute_result" } ], "source": [ - "filter_dgp = \"truncated_normal\" # filter by DGP\n", - "results_df = preprocess_results_df(all_results_df.loc[all_results_df[\"algorithm\"].isin([\"kl_dro_bas\", \"kl_pp\", \"kl_bdro\"])], filter_dgp)\n", + "filter_dgp = \"normal\" # filter by DGP\n", + "results_df = preprocess_results_df(all_results_df, filter_dgp)\n", "# results_df = preprocess_results_df(all_results_df, filter_dgp)\n", "results_df.head(3)" ] }, { "cell_type": "code", - "execution_count": 4, + "execution_count": 12, "metadata": {}, "outputs": [], "source": [ @@ -873,16 +312,16 @@ }, { "cell_type": "code", - "execution_count": 5, + "execution_count": 13, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ - "/Users/patrick/Projects/mis-dro-code/mis_dro/plot.py:214: FutureWarning: The provided callable is currently using SeriesGroupBy.mean. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"mean\" instead.\n", + "/Users/patrick/Projects/mis-dro-code/mis_dro/results.py:66: FutureWarning: The provided callable is currently using SeriesGroupBy.mean. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"mean\" instead.\n", " agg_df = gb.agg(\n", - "/Users/patrick/Projects/mis-dro-code/mis_dro/plot.py:214: FutureWarning: The provided callable is currently using SeriesGroupBy.std. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"std\" instead.\n", + "/Users/patrick/Projects/mis-dro-code/mis_dro/results.py:66: FutureWarning: The provided callable is currently using SeriesGroupBy.std. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"std\" instead.\n", " agg_df = gb.agg(\n" ] }, @@ -914,6 +353,7 @@ " \n", " out_of_sample_mean\n", " out_of_sample_var\n", + " out_of_sample_std\n", " sum_of_in_group_var\n", " var_of_in_group_mean\n", " mean_solve_time\n", @@ -936,74 +376,80 @@ " \n", " \n", " \n", + " \n", " \n", " \n", " \n", " \n", " kl_bdro\n", - " truncated_normal\n", + " normal\n", " 0.001\n", " bayes\n", " 25\n", " 20\n", - " 33.183397\n", - " 690.722569\n", - " 663.527418\n", - " 27.195151\n", - " 0.067703\n", - " 0.011895\n", - " 0.000119\n", - " 0.000008\n", + " 39.210286\n", + " 923.900937\n", + " 30.395739\n", + " 890.666964\n", + " 33.233973\n", + " 0.068447\n", + " 1.201670e-02\n", + " 0.000124\n", + " 1.013000e-05\n", " \n", " \n", " 100\n", " 20\n", - " 32.479560\n", - " 655.081832\n", - " 635.906347\n", - " 19.175485\n", - " 0.145540\n", - " 0.018745\n", - " 0.000149\n", - " 0.000008\n", + " 37.910862\n", + " 864.472885\n", + " 29.401920\n", + " 841.225912\n", + " 23.246974\n", + " 0.145043\n", + " 1.853992e-02\n", + " 0.000152\n", + " 8.410029e-06\n", " \n", " \n", " 900\n", " 20\n", - " 32.238666\n", - " 641.350995\n", - " 625.535744\n", - " 15.815251\n", - " 0.716681\n", - " 0.030649\n", - " 0.000296\n", - " 0.000013\n", + " 37.555099\n", + " 838.254272\n", + " 28.952621\n", + " 817.627933\n", + " 20.626339\n", + " 0.727444\n", + " 3.144634e-02\n", + " 0.000292\n", + " 1.277644e-05\n", " \n", " \n", " 0.002\n", " bayes\n", " 25\n", " 20\n", - " 33.162919\n", - " 691.583013\n", - " 664.586005\n", - " 26.997008\n", - " 0.068123\n", - " 0.011182\n", - " 0.000120\n", - " 0.000004\n", + " 39.171374\n", + " 924.516225\n", + " 30.405858\n", + " 891.584082\n", + " 32.932143\n", + " 0.070016\n", + " 1.147892e-02\n", + " 0.000124\n", + " 5.443806e-06\n", " \n", " \n", " 100\n", " 20\n", - " 32.491418\n", - " 653.271270\n", - " 634.143757\n", - " 19.127512\n", - " 0.147815\n", - " 0.020667\n", - " 0.000152\n", - " 0.000010\n", + " 37.926581\n", + " 862.524447\n", + " 29.368767\n", + " 839.227196\n", + " 23.297251\n", + " 0.147706\n", + " 2.043214e-02\n", + " 0.000153\n", + " 1.133777e-05\n", " \n", " \n", " ...\n", @@ -1020,195 +466,221 @@ " ...\n", " ...\n", " ...\n", + " ...\n", " \n", " \n", - " kl_pp\n", - " truncated_normal\n", - " 2.500\n", - " bayes\n", - " 100\n", + " wasserstein_empirical\n", + " normal\n", + " 40.000\n", + " empirical\n", + " 20\n", " 20\n", - " 39.030519\n", - " 568.064303\n", - " 474.987064\n", - " 93.077239\n", - " 0.034726\n", - " 0.001168\n", - " 0.000142\n", - " 0.000006\n", + " 79.690249\n", + " 940.611243\n", + " 30.669386\n", + " 849.943942\n", + " 90.667301\n", + " 0.000019\n", + " 1.316269e-06\n", + " 0.000010\n", + " 9.645085e-07\n", " \n", " \n", - " 900\n", + " 45.000\n", + " empirical\n", + " 20\n", " 20\n", - " 43.380106\n", - " 564.164230\n", - " 449.611358\n", - " 114.552872\n", - " 0.382384\n", - " 0.012819\n", - " 0.000182\n", - " 0.000011\n", + " 87.240542\n", + " 951.770005\n", + " 30.850770\n", + " 859.214885\n", + " 92.555120\n", + " 0.000019\n", + " 8.139540e-07\n", + " 0.000010\n", + " 7.900280e-07\n", " \n", " \n", - " 3.000\n", - " bayes\n", - " 25\n", + " 50.000\n", + " empirical\n", " 20\n", - " 35.567185\n", - " 647.054885\n", - " 591.312568\n", - " 55.742317\n", - " 0.022117\n", - " 0.000698\n", - " 0.000135\n", - " 0.000006\n", + " 20\n", + " 94.844418\n", + " 957.783377\n", + " 30.948075\n", + " 864.175882\n", + " 93.607495\n", + " 0.000019\n", + " 8.873979e-07\n", + " 0.000010\n", + " 8.213191e-07\n", " \n", " \n", - " 100\n", + " 55.000\n", + " empirical\n", + " 20\n", " 20\n", - " 39.129655\n", - " 572.370428\n", - " 476.941854\n", - " 95.428574\n", - " 0.036588\n", - " 0.003450\n", - " 0.000149\n", + " 102.471480\n", + " 961.563197\n", + " 31.009082\n", + " 867.366077\n", + " 94.197121\n", + " 0.000019\n", + " 1.273644e-06\n", " 0.000010\n", + " 8.370647e-07\n", " \n", " \n", - " 900\n", + " 60.000\n", + " empirical\n", " 20\n", - " 44.322418\n", - " 580.245499\n", - " 452.675654\n", - " 127.569845\n", - " 0.388065\n", - " 0.013021\n", - " 0.000190\n", - " 0.000012\n", + " 20\n", + " 110.110576\n", + " 964.076645\n", + " 31.049584\n", + " 869.523912\n", + " 94.552733\n", + " 0.000019\n", + " 9.168452e-07\n", + " 0.000010\n", + " 6.906504e-07\n", " \n", " \n", "\n", - "

231 rows × 8 columns

\n", + "

289 rows × 9 columns

\n", "" ], "text/plain": [ - " out_of_sample_mean \\\n", - "algorithm dgp epsilon inference num_total_samples num_observations \n", - "kl_bdro truncated_normal 0.001 bayes 25 20 33.183397 \n", - " 100 20 32.479560 \n", - " 900 20 32.238666 \n", - " 0.002 bayes 25 20 33.162919 \n", - " 100 20 32.491418 \n", - "... ... \n", - "kl_pp truncated_normal 2.500 bayes 100 20 39.030519 \n", - " 900 20 43.380106 \n", - " 3.000 bayes 25 20 35.567185 \n", - " 100 20 39.129655 \n", - " 900 20 44.322418 \n", + " out_of_sample_mean \\\n", + "algorithm dgp epsilon inference num_total_samples num_observations \n", + "kl_bdro normal 0.001 bayes 25 20 39.210286 \n", + " 100 20 37.910862 \n", + " 900 20 37.555099 \n", + " 0.002 bayes 25 20 39.171374 \n", + " 100 20 37.926581 \n", + "... ... \n", + "wasserstein_empirical normal 40.000 empirical 20 20 79.690249 \n", + " 45.000 empirical 20 20 87.240542 \n", + " 50.000 empirical 20 20 94.844418 \n", + " 55.000 empirical 20 20 102.471480 \n", + " 60.000 empirical 20 20 110.110576 \n", "\n", - " out_of_sample_var \\\n", - "algorithm dgp epsilon inference num_total_samples num_observations \n", - "kl_bdro truncated_normal 0.001 bayes 25 20 690.722569 \n", - " 100 20 655.081832 \n", - " 900 20 641.350995 \n", - " 0.002 bayes 25 20 691.583013 \n", - " 100 20 653.271270 \n", - "... ... \n", - "kl_pp truncated_normal 2.500 bayes 100 20 568.064303 \n", - " 900 20 564.164230 \n", - " 3.000 bayes 25 20 647.054885 \n", - " 100 20 572.370428 \n", - " 900 20 580.245499 \n", + " out_of_sample_var \\\n", + "algorithm dgp epsilon inference num_total_samples num_observations \n", + "kl_bdro normal 0.001 bayes 25 20 923.900937 \n", + " 100 20 864.472885 \n", + " 900 20 838.254272 \n", + " 0.002 bayes 25 20 924.516225 \n", + " 100 20 862.524447 \n", + "... ... \n", + "wasserstein_empirical normal 40.000 empirical 20 20 940.611243 \n", + " 45.000 empirical 20 20 951.770005 \n", + " 50.000 empirical 20 20 957.783377 \n", + " 55.000 empirical 20 20 961.563197 \n", + " 60.000 empirical 20 20 964.076645 \n", "\n", - " sum_of_in_group_var \\\n", - "algorithm dgp epsilon inference num_total_samples num_observations \n", - "kl_bdro truncated_normal 0.001 bayes 25 20 663.527418 \n", - " 100 20 635.906347 \n", - " 900 20 625.535744 \n", - " 0.002 bayes 25 20 664.586005 \n", - " 100 20 634.143757 \n", + " out_of_sample_std \\\n", + "algorithm dgp epsilon inference num_total_samples num_observations \n", + "kl_bdro normal 0.001 bayes 25 20 30.395739 \n", + " 100 20 29.401920 \n", + " 900 20 28.952621 \n", + " 0.002 bayes 25 20 30.405858 \n", + " 100 20 29.368767 \n", "... ... \n", - "kl_pp truncated_normal 2.500 bayes 100 20 474.987064 \n", - " 900 20 449.611358 \n", - " 3.000 bayes 25 20 591.312568 \n", - " 100 20 476.941854 \n", - " 900 20 452.675654 \n", + "wasserstein_empirical normal 40.000 empirical 20 20 30.669386 \n", + " 45.000 empirical 20 20 30.850770 \n", + " 50.000 empirical 20 20 30.948075 \n", + " 55.000 empirical 20 20 31.009082 \n", + " 60.000 empirical 20 20 31.049584 \n", "\n", - " var_of_in_group_mean \\\n", - "algorithm dgp epsilon inference num_total_samples num_observations \n", - "kl_bdro truncated_normal 0.001 bayes 25 20 27.195151 \n", - " 100 20 19.175485 \n", - " 900 20 15.815251 \n", - " 0.002 bayes 25 20 26.997008 \n", - " 100 20 19.127512 \n", - "... ... \n", - "kl_pp truncated_normal 2.500 bayes 100 20 93.077239 \n", - " 900 20 114.552872 \n", - " 3.000 bayes 25 20 55.742317 \n", - " 100 20 95.428574 \n", - " 900 20 127.569845 \n", + " sum_of_in_group_var \\\n", + "algorithm dgp epsilon inference num_total_samples num_observations \n", + "kl_bdro normal 0.001 bayes 25 20 890.666964 \n", + " 100 20 841.225912 \n", + " 900 20 817.627933 \n", + " 0.002 bayes 25 20 891.584082 \n", + " 100 20 839.227196 \n", + "... ... \n", + "wasserstein_empirical normal 40.000 empirical 20 20 849.943942 \n", + " 45.000 empirical 20 20 859.214885 \n", + " 50.000 empirical 20 20 864.175882 \n", + " 55.000 empirical 20 20 867.366077 \n", + " 60.000 empirical 20 20 869.523912 \n", "\n", - " mean_solve_time \\\n", - "algorithm dgp epsilon inference num_total_samples num_observations \n", - "kl_bdro truncated_normal 0.001 bayes 25 20 0.067703 \n", - " 100 20 0.145540 \n", - " 900 20 0.716681 \n", - " 0.002 bayes 25 20 0.068123 \n", - " 100 20 0.147815 \n", - "... ... \n", - "kl_pp truncated_normal 2.500 bayes 100 20 0.034726 \n", - " 900 20 0.382384 \n", - " 3.000 bayes 25 20 0.022117 \n", - " 100 20 0.036588 \n", - " 900 20 0.388065 \n", + " var_of_in_group_mean \\\n", + "algorithm dgp epsilon inference num_total_samples num_observations \n", + "kl_bdro normal 0.001 bayes 25 20 33.233973 \n", + " 100 20 23.246974 \n", + " 900 20 20.626339 \n", + " 0.002 bayes 25 20 32.932143 \n", + " 100 20 23.297251 \n", + "... ... \n", + "wasserstein_empirical normal 40.000 empirical 20 20 90.667301 \n", + " 45.000 empirical 20 20 92.555120 \n", + " 50.000 empirical 20 20 93.607495 \n", + " 55.000 empirical 20 20 94.197121 \n", + " 60.000 empirical 20 20 94.552733 \n", "\n", - " std_solve_time \\\n", - "algorithm dgp epsilon inference num_total_samples num_observations \n", - "kl_bdro truncated_normal 0.001 bayes 25 20 0.011895 \n", - " 100 20 0.018745 \n", - " 900 20 0.030649 \n", - " 0.002 bayes 25 20 0.011182 \n", - " 100 20 0.020667 \n", - "... ... \n", - "kl_pp truncated_normal 2.500 bayes 100 20 0.001168 \n", - " 900 20 0.012819 \n", - " 3.000 bayes 25 20 0.000698 \n", - " 100 20 0.003450 \n", - " 900 20 0.013021 \n", + " mean_solve_time \\\n", + "algorithm dgp epsilon inference num_total_samples num_observations \n", + "kl_bdro normal 0.001 bayes 25 20 0.068447 \n", + " 100 20 0.145043 \n", + " 900 20 0.727444 \n", + " 0.002 bayes 25 20 0.070016 \n", + " 100 20 0.147706 \n", + "... ... \n", + "wasserstein_empirical normal 40.000 empirical 20 20 0.000019 \n", + " 45.000 empirical 20 20 0.000019 \n", + " 50.000 empirical 20 20 0.000019 \n", + " 55.000 empirical 20 20 0.000019 \n", + " 60.000 empirical 20 20 0.000019 \n", "\n", - " mean_sample_time \\\n", - "algorithm dgp epsilon inference num_total_samples num_observations \n", - "kl_bdro truncated_normal 0.001 bayes 25 20 0.000119 \n", - " 100 20 0.000149 \n", - " 900 20 0.000296 \n", - " 0.002 bayes 25 20 0.000120 \n", - " 100 20 0.000152 \n", + " std_solve_time \\\n", + "algorithm dgp epsilon inference num_total_samples num_observations \n", + "kl_bdro normal 0.001 bayes 25 20 1.201670e-02 \n", + " 100 20 1.853992e-02 \n", + " 900 20 3.144634e-02 \n", + " 0.002 bayes 25 20 1.147892e-02 \n", + " 100 20 2.043214e-02 \n", "... ... \n", - "kl_pp truncated_normal 2.500 bayes 100 20 0.000142 \n", - " 900 20 0.000182 \n", - " 3.000 bayes 25 20 0.000135 \n", - " 100 20 0.000149 \n", - " 900 20 0.000190 \n", + "wasserstein_empirical normal 40.000 empirical 20 20 1.316269e-06 \n", + " 45.000 empirical 20 20 8.139540e-07 \n", + " 50.000 empirical 20 20 8.873979e-07 \n", + " 55.000 empirical 20 20 1.273644e-06 \n", + " 60.000 empirical 20 20 9.168452e-07 \n", "\n", - " std_sample_time \n", - "algorithm dgp epsilon inference num_total_samples num_observations \n", - "kl_bdro truncated_normal 0.001 bayes 25 20 0.000008 \n", - " 100 20 0.000008 \n", - " 900 20 0.000013 \n", - " 0.002 bayes 25 20 0.000004 \n", - " 100 20 0.000010 \n", - "... ... \n", - "kl_pp truncated_normal 2.500 bayes 100 20 0.000006 \n", - " 900 20 0.000011 \n", - " 3.000 bayes 25 20 0.000006 \n", - " 100 20 0.000010 \n", - " 900 20 0.000012 \n", + " mean_sample_time \\\n", + "algorithm dgp epsilon inference num_total_samples num_observations \n", + "kl_bdro normal 0.001 bayes 25 20 0.000124 \n", + " 100 20 0.000152 \n", + " 900 20 0.000292 \n", + " 0.002 bayes 25 20 0.000124 \n", + " 100 20 0.000153 \n", + "... ... \n", + "wasserstein_empirical normal 40.000 empirical 20 20 0.000010 \n", + " 45.000 empirical 20 20 0.000010 \n", + " 50.000 empirical 20 20 0.000010 \n", + " 55.000 empirical 20 20 0.000010 \n", + " 60.000 empirical 20 20 0.000010 \n", + "\n", + " std_sample_time \n", + "algorithm dgp epsilon inference num_total_samples num_observations \n", + "kl_bdro normal 0.001 bayes 25 20 1.013000e-05 \n", + " 100 20 8.410029e-06 \n", + " 900 20 1.277644e-05 \n", + " 0.002 bayes 25 20 5.443806e-06 \n", + " 100 20 1.133777e-05 \n", + "... ... \n", + "wasserstein_empirical normal 40.000 empirical 20 20 9.645085e-07 \n", + " 45.000 empirical 20 20 7.900280e-07 \n", + " 50.000 empirical 20 20 8.213191e-07 \n", + " 55.000 empirical 20 20 8.370647e-07 \n", + " 60.000 empirical 20 20 6.906504e-07 \n", "\n", - "[231 rows x 8 columns]" + "[289 rows x 9 columns]" ] }, - "execution_count": 5, + "execution_count": 13, "metadata": {}, "output_type": "execute_result" } @@ -1229,7 +701,7 @@ }, { "cell_type": "code", - "execution_count": 6, + "execution_count": 14, "metadata": {}, "outputs": [ { @@ -1243,7 +715,7 @@ }, { "data": { - "image/png": "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", + "image/png": "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", "text/plain": [ "
" ] @@ -1253,7 +725,7 @@ } ], "source": [ - "total_samples_list = agg_df.loc[~(agg_df.index.get_level_values(\"algorithm\") == \"kl_empirical\")].index.get_level_values(\"num_total_samples\").unique().tolist()\n", + "total_samples_list = agg_df.loc[~(agg_df.index.get_level_values(\"algorithm\").isin((\"kl_empirical\", \"wasserstein_empirical\")))].index.get_level_values(\"num_total_samples\").unique().tolist()\n", "# num_observations_list = agg_df.index.get_level_values(\"num_observations\").unique().tolist()\n", "dgp_list = agg_df.index.get_level_values(\"dgp\").unique().tolist()\n", "\n", @@ -1261,6 +733,7 @@ "ncols = len(total_samples_list)\n", "# ncols = len(num_observations_list)\n", "trim_epsilon = 1.0\n", + "wass_trim_epsilon = 10.0\n", "\n", "fig, axes = plt.subplots(nrows=nrows, ncols=ncols, sharex=True, sharey=True, figsize=(5*ncols, nrows*4))\n", "fig.subplots_adjust(wspace=0.05)\n", @@ -1270,9 +743,10 @@ " # for j, num_observations in enumerate(num_observations_list):\n", "\n", " # NOTE empirical KL doesn't sample, so we plot it before filtering by total_samples\n", - " # mean_variance_plot(axes[j], agg_df.loc[\"kl_empirical\", dgp, :trim_epsilon, \"empirical\", :, :], **algorithm_inference_style(\"kl_empirical\", \"empirical\", label_inference=False), special_epsilons=[0.3], offset=0.1)\n", + " mean_variance_plot(axes[j], agg_df.loc[\"kl_empirical\", dgp, :trim_epsilon, \"empirical\", :, :], **algorithm_inference_style(\"kl_empirical\", \"empirical\", label_inference=False), special_epsilons=[0.3], offset=0.1)\n", " # mean_variance_plot(axes[j], agg_df.loc[\"kl_empirical\", dgp, :trim_epsilon, \"empirical\", num_observations, :], **algorithm_inference_style(\"kl_empirical\", \"empirical\", label_inference=False), special_epsilons=[0.3], offset=0.1)\n", - " \n", + " mean_variance_plot(axes[j], agg_df.loc[\"wasserstein_empirical\", dgp, :wass_trim_epsilon, \"empirical\", :, :], **algorithm_inference_style(\"wasserstein_empirical\", \"empirical\", label_inference=False), special_epsilons=[0.3], offset=0.1)\n", + "\n", "\n", " # filter by the total samples and DGP\n", " axis_df = agg_df.loc[:, dgp, :, :, total_samples, :]\n", @@ -1285,26 +759,32 @@ " print(\"All BDRO points are Pareto dominated for M =\", total_samples, \"?\", not df[\"is_pareto_front\"].any())\n", "\n", "\n", - " for k in range(j, len(total_samples_list)):\n", - " # for k in range(j, len(num_observations_list)):\n", - " alpha = 1.0\n", - " is_labelled=True\n", - " if j != k:\n", - " alpha = 0.2\n", - " is_labelled = False\n", - " mean_variance_plot(axes[k], df, **algorithm_inference_style(algorithm, inference, label_inference=False), offset=0.0, is_labelled=is_labelled, alpha=alpha)\n", + " # for k in range(j, len(total_samples_list)):\n", + " # # for k in range(j, len(num_observations_list)):\n", + " # alpha = 1.0\n", + " # is_labelled=True\n", + " # if j != k:\n", + " # alpha = 0.2\n", + " # is_labelled = False\n", + " # mean_variance_plot(axes[k], df, **algorithm_inference_style(algorithm, inference, label_inference=False), offset=0.0, is_labelled=is_labelled, alpha=alpha)\n", + " alpha = 1.0\n", + " is_labelled=True\n", + " mean_variance_plot(axes[j], df, **algorithm_inference_style(algorithm, inference, label_inference=False), offset=0.0, is_labelled=is_labelled, alpha=alpha)\n", + "\n", " if j == 0:\n", + " axes[j].set_ylabel(\"Out-of-sample mean, $m(\\epsilon)$\")\n", + " if j == 2:\n", " handles, labels = axes[j].get_legend_handles_labels()\n", - " # algorithm_order = [AlgorithmName.kl_dro_bas.value, AlgorithmName.kl_pp.value, AlgorithmName.kl_bdro.value, AlgorithmName.kl_empirical.value]\n", - " algorithm_order = [AlgorithmName.kl_dro_bas.value, AlgorithmName.kl_pp.value, AlgorithmName.kl_bdro.value]\n", + " # algorithm_order = [AlgorithmName.kl_dro_bas.value, AlgorithmName.kl_pp.value, AlgorithmName.kl_bdro.value]\n", + " algorithm_order = [AlgorithmName.kl_dro_bas.value, AlgorithmName.kl_pp.value, AlgorithmName.kl_bdro.value, AlgorithmName.kl_empirical.value, AlgorithmName.wasserstein_empirical.value]\n", " order = [list(labels).index(a) for a in algorithm_order]\n", " axes[j].legend([handles[idx] for idx in order],[labels[idx] for idx in order])\n", - " axes[j].set_ylabel(\"Out-of-sample mean, $m(\\epsilon)$\")\n", " axes[j].set_title(\"$M$\" + f\"={total_samples} with {NiceNameDGP[filter_dgp]}\")\n", " # axes[j].set_title(\"$n$\" + f\"={num_observations} with {NiceNameDGP[filter_dgp]}\")\n", " axes[j].set_xlabel(\"Out-of-sample variance, $v(\\epsilon)$\")\n", "\n", - "# fig.savefig(f\"/Users/patrick/Experiments/misdro/2025_01_21_paper_bas_experiments/{experiment_name}_{filter_dgp}_empirical.pdf\", bbox_inches=\"tight\")" + "# fig.savefig(f\"/Users/patrick/Experiments/misdro/2025_01_21_paper_bas_experiments/{experiment_name}_{filter_dgp}_empirical.pdf\", bbox_inches=\"tight\")\n", + "fig.savefig(f\"/Users/patrick/Experiments/misdro/2025_01_21_paper_bas_experiments/{experiment_name}_{filter_dgp}_empirical_wasserstein.pdf\", bbox_inches=\"tight\")" ] }, { @@ -1316,7 +796,7 @@ }, { "cell_type": "code", - "execution_count": 7, + "execution_count": 8, "metadata": {}, "outputs": [ { @@ -1329,7 +809,8 @@ " & & kl_dro_bas & kl_pp & kl_bdro & kl_dro_bas & kl_pp & kl_bdro \\\\\n", "dgp & num_total_samples & & & & & & \\\\\n", "\\midrule\n", - "\\multirow[t]{3}{*}{truncated_normal} & 25 & 0.024 (0.003) & 0.024 (0.003) & 0.067 (0.012) & 0.070 (0.006) & 0.138 (0.008) & 0.121 (0.009) \\\\\n", + "\\multirow[t]{4}{*}{truncated_normal} & 20 & NaN & NaN & NaN & NaN & NaN & NaN \\\\\n", + " & 25 & 0.024 (0.003) & 0.024 (0.003) & 0.067 (0.012) & 0.070 (0.006) & 0.138 (0.008) & 0.121 (0.009) \\\\\n", " & 100 & 0.035 (0.003) & 0.035 (0.003) & 0.145 (0.020) & 0.072 (0.006) & 0.142 (0.009) & 0.156 (0.013) \\\\\n", " & 900 & 0.398 (0.020) & 0.406 (0.021) & 0.683 (0.033) & 0.090 (0.008) & 0.185 (0.012) & 0.292 (0.025) \\\\\n", "\\cline{1-8}\n", @@ -1342,9 +823,9 @@ "name": "stderr", "output_type": "stream", "text": [ - "/Users/patrick/Projects/mis-dro-code/mis_dro/plot.py:214: FutureWarning: The provided callable is currently using SeriesGroupBy.mean. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"mean\" instead.\n", + "/Users/patrick/Projects/mis-dro-code/mis_dro/results.py:66: FutureWarning: The provided callable is currently using SeriesGroupBy.mean. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"mean\" instead.\n", " agg_df = gb.agg(\n", - "/Users/patrick/Projects/mis-dro-code/mis_dro/plot.py:214: FutureWarning: The provided callable is currently using SeriesGroupBy.std. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"std\" instead.\n", + "/Users/patrick/Projects/mis-dro-code/mis_dro/results.py:66: FutureWarning: The provided callable is currently using SeriesGroupBy.std. In a future version of pandas, the provided callable will be used directly. To keep current behavior pass the string \"std\" instead.\n", " agg_df = gb.agg(\n" ] } @@ -1390,7 +871,7 @@ }, { "cell_type": "code", - "execution_count": 8, + "execution_count": 9, "metadata": {}, "outputs": [ { @@ -1408,7 +889,7 @@ "\u001b[0;31mKeyError\u001b[0m: 25", "\nThe above exception was the direct cause of the following exception:\n", "\u001b[0;31mKeyError\u001b[0m Traceback (most recent call last)", - "Cell \u001b[0;32mIn[8], line 21\u001b[0m\n\u001b[1;32m 19\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m i, var_col \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28menumerate\u001b[39m([\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mout_of_sample_var\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124msum_of_in_group_var\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mvar_of_in_group_mean\u001b[39m\u001b[38;5;124m\"\u001b[39m]):\n\u001b[1;32m 20\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m j, total_samples \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28menumerate\u001b[39m(total_samples_list):\n\u001b[0;32m---> 21\u001b[0m axis_df \u001b[38;5;241m=\u001b[39m \u001b[43magg_df\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mloc\u001b[49m\u001b[43m[\u001b[49m\u001b[43m:\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mdgp\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43m:\u001b[49m\u001b[43mtrim_epsilon\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mtotal_samples\u001b[49m\u001b[43m]\u001b[49m\n\u001b[1;32m 22\u001b[0m axis_df\u001b[38;5;241m.\u001b[39mloc[:, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mis_pareto_front\u001b[39m\u001b[38;5;124m\"\u001b[39m] \u001b[38;5;241m=\u001b[39m is_minimise_pareto_front(axis_df[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mout_of_sample_var\u001b[39m\u001b[38;5;124m\"\u001b[39m]\u001b[38;5;241m.\u001b[39mvalues, axis_df[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mout_of_sample_mean\u001b[39m\u001b[38;5;124m\"\u001b[39m]\u001b[38;5;241m.\u001b[39mvalues)\n\u001b[1;32m 23\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m algorithm \u001b[38;5;129;01min\u001b[39;00m axis_df\u001b[38;5;241m.\u001b[39mindex\u001b[38;5;241m.\u001b[39mget_level_values(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124malgorithm\u001b[39m\u001b[38;5;124m\"\u001b[39m)\u001b[38;5;241m.\u001b[39munique():\n", + "Cell \u001b[0;32mIn[9], line 21\u001b[0m\n\u001b[1;32m 19\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m i, var_col \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28menumerate\u001b[39m([\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mout_of_sample_var\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124msum_of_in_group_var\u001b[39m\u001b[38;5;124m\"\u001b[39m, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mvar_of_in_group_mean\u001b[39m\u001b[38;5;124m\"\u001b[39m]):\n\u001b[1;32m 20\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m j, total_samples \u001b[38;5;129;01min\u001b[39;00m \u001b[38;5;28menumerate\u001b[39m(total_samples_list):\n\u001b[0;32m---> 21\u001b[0m axis_df \u001b[38;5;241m=\u001b[39m \u001b[43magg_df\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mloc\u001b[49m\u001b[43m[\u001b[49m\u001b[43m:\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mdgp\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43m:\u001b[49m\u001b[43mtrim_epsilon\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mtotal_samples\u001b[49m\u001b[43m]\u001b[49m\n\u001b[1;32m 22\u001b[0m axis_df\u001b[38;5;241m.\u001b[39mloc[:, \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mis_pareto_front\u001b[39m\u001b[38;5;124m\"\u001b[39m] \u001b[38;5;241m=\u001b[39m is_minimise_pareto_front(axis_df[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mout_of_sample_var\u001b[39m\u001b[38;5;124m\"\u001b[39m]\u001b[38;5;241m.\u001b[39mvalues, axis_df[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mout_of_sample_mean\u001b[39m\u001b[38;5;124m\"\u001b[39m]\u001b[38;5;241m.\u001b[39mvalues)\n\u001b[1;32m 23\u001b[0m \u001b[38;5;28;01mfor\u001b[39;00m algorithm \u001b[38;5;129;01min\u001b[39;00m axis_df\u001b[38;5;241m.\u001b[39mindex\u001b[38;5;241m.\u001b[39mget_level_values(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124malgorithm\u001b[39m\u001b[38;5;124m\"\u001b[39m)\u001b[38;5;241m.\u001b[39munique():\n", "File \u001b[0;32m/opt/anaconda3/envs/mis-dro/lib/python3.11/site-packages/pandas/core/indexing.py:1184\u001b[0m, in \u001b[0;36m_LocationIndexer.__getitem__\u001b[0;34m(self, key)\u001b[0m\n\u001b[1;32m 1182\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_is_scalar_access(key):\n\u001b[1;32m 1183\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mobj\u001b[38;5;241m.\u001b[39m_get_value(\u001b[38;5;241m*\u001b[39mkey, takeable\u001b[38;5;241m=\u001b[39m\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_takeable)\n\u001b[0;32m-> 1184\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_getitem_tuple\u001b[49m\u001b[43m(\u001b[49m\u001b[43mkey\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 1185\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m 1186\u001b[0m \u001b[38;5;66;03m# we by definition only have the 0th axis\u001b[39;00m\n\u001b[1;32m 1187\u001b[0m axis \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39maxis \u001b[38;5;129;01mor\u001b[39;00m \u001b[38;5;241m0\u001b[39m\n", "File \u001b[0;32m/opt/anaconda3/envs/mis-dro/lib/python3.11/site-packages/pandas/core/indexing.py:1368\u001b[0m, in \u001b[0;36m_LocIndexer._getitem_tuple\u001b[0;34m(self, tup)\u001b[0m\n\u001b[1;32m 1366\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m suppress(IndexingError):\n\u001b[1;32m 1367\u001b[0m tup \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_expand_ellipsis(tup)\n\u001b[0;32m-> 1368\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_getitem_lowerdim\u001b[49m\u001b[43m(\u001b[49m\u001b[43mtup\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 1370\u001b[0m \u001b[38;5;66;03m# no multi-index, so validate all of the indexers\u001b[39;00m\n\u001b[1;32m 1371\u001b[0m tup \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_validate_tuple_indexer(tup)\n", "File \u001b[0;32m/opt/anaconda3/envs/mis-dro/lib/python3.11/site-packages/pandas/core/indexing.py:1041\u001b[0m, in \u001b[0;36m_LocationIndexer._getitem_lowerdim\u001b[0;34m(self, tup)\u001b[0m\n\u001b[1;32m 1039\u001b[0m \u001b[38;5;66;03m# we may have a nested tuples indexer here\u001b[39;00m\n\u001b[1;32m 1040\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_is_nested_tuple_indexer(tup):\n\u001b[0;32m-> 1041\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_getitem_nested_tuple\u001b[49m\u001b[43m(\u001b[49m\u001b[43mtup\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 1043\u001b[0m \u001b[38;5;66;03m# we maybe be using a tuple to represent multiple dimensions here\u001b[39;00m\n\u001b[1;32m 1044\u001b[0m ax0 \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mobj\u001b[38;5;241m.\u001b[39m_get_axis(\u001b[38;5;241m0\u001b[39m)\n", @@ -1423,9 +904,9 @@ }, { "data": { - "image/png": "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", + "image/png": "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", "text/plain": [ - "
" + "
" ] }, "metadata": {}, @@ -1478,7 +959,7 @@ }, { "cell_type": "code", - "execution_count": 34, + "execution_count": null, "metadata": {}, "outputs": [ {