Skip to content

image_display_example.md

Latest commit

 

History

History
171 lines (135 loc) · 5.32 KB

image_display_example.md

File metadata and controls

171 lines (135 loc) · 5.32 KB

RMCP Visual Analytics Example

New in v0.3.7: This example demonstrates how RMCP now displays professional-quality plots and visualizations directly in Claude conversations, revolutionizing data analysis workflows.

🎯 Revolutionary Visual Analytics

Direct Image Display in Claude

RMCP visualization tools now return both comprehensive statistical analysis and publication-quality images directly in Claude:

# When you ask Claude:
"Create a correlation heatmap of my sales, marketing, and customer satisfaction data"

# RMCP responds with:
# 1. 📊 Interactive heatmap displayed inline with color-coded correlation strengths
# 2. 📋 Statistical analysis: correlation matrix with exact values and significance tests
# 3. 💡 Insights: "Strong positive correlation (r=0.89) between marketing and sales"
# 4. 🎨 Professional ggplot2 styling ready for presentations

🎨 Enhanced Visualization Tools (All 6 Support Inline Display)

🔥 correlation_heatmap: Color-coded correlation matrices with statistical significance testing

  • Perfect for: Exploring relationships between multiple variables
  • Visual: Color intensity shows correlation strength (-1 to +1)
  • Analysis: p-values, confidence intervals, sample sizes

📈 scatter_plot: Interactive scatter plots with trend lines and grouping

  • Perfect for: Regression analysis, outlier detection, group comparisons
  • Visual: Points, trend lines, confidence bands, group colors
  • Analysis: Correlation coefficients, R², regression equations

📊 histogram: Distribution analysis with density overlays

  • Perfect for: Understanding data distributions, checking normality
  • Visual: Bars with density curves, group overlays
  • Analysis: Mean, median, skewness, kurtosis statistics

📦 boxplot: Quartile analysis with outlier detection

  • Perfect for: Comparing distributions, finding outliers
  • Visual: Boxes, whiskers, outlier points, group comparisons
  • Analysis: Quartiles, IQR, outlier counts, group statistics

⏱️ time_series_plot: Temporal analysis with trend forecasting

  • Perfect for: Time series analysis, trend identification
  • Visual: Lines, points, smooth trends, confidence bands
  • Analysis: Trend statistics, seasonal patterns, forecasts

🔍 regression_plot: Comprehensive diagnostic plots (4-panel)

  • Perfect for: Model validation, assumption checking
  • Visual: Residuals vs fitted, Q-Q plots, scale-location, leverage
  • Analysis: Model diagnostics, outliers, influential points

Usage Examples

1. Correlation Heatmap (no file needed)

{
  "tool": "correlation_heatmap",
  "arguments": {
    "data": {
      "sales": [100, 150, 200, 250, 300],
      "marketing": [10, 15, 25, 30, 40],
      "temperature": [20, 25, 30, 35, 40]
    },
    "method": "pearson",
    "title": "Sales Correlation Analysis"
  }
}

Returns:

  • Text: Correlation matrix with values, statistics
  • Image: Color-coded heatmap displayed directly in Claude

2. Scatter Plot with Grouping

{
  "tool": "scatter_plot",
  "arguments": {
    "data": {
      "x": [1, 2, 3, 4, 5, 6, 7, 8],
      "y": [2, 4, 3, 6, 5, 8, 7, 10],
      "group": ["A", "A", "B", "B", "A", "A", "B", "B"]
    },
    "x": "x",
    "y": "y",
    "group": "group",
    "title": "Sales vs Marketing by Region"
  }
}

Returns:

  • Text: Correlation coefficient, data points count
  • Image: Scatter plot with color-coded groups and trend lines

Optional File Saving

You can still save plots to files if needed:

{
  "tool": "correlation_heatmap",
  "arguments": {
    "data": {...},
    "file_path": "/path/to/save/heatmap.png",
    "return_image": true
  }
}

This saves the plot to a file and displays it inline in Claude.

Technical Details

Image Format

  • Format: PNG images with white background
  • Encoding: Base64 for transmission
  • Resolution: Configurable (default 800x600 pixels)
  • Quality: 100 DPI for crisp display

MCP Content Response

Tools now return multiple content types:

{
  "content": [
    {
      "type": "text",
      "text": "{\"correlation_matrix\": [[1.0, 0.95], [0.95, 1.0]], ...}"
    },
    {
      "type": "image",
      "data": "iVBORw0KGgoAAAANSUhEUgAAA...",
      "mimeType": "image/png"
    }
  ]
}

Configuration Options

All visualization tools support these parameters:

{
  "return_image": true,     // Enable/disable inline images (default: true)
  "file_path": "plot.png",  // Optional: also save to file
  "width": 800,             // Image width in pixels
  "height": 600             // Image height in pixels
}

Benefits

  1. Immediate Visual Feedback: See plots instantly without file management
  2. Streamlined Workflow: Analysis and visualization in one conversation
  3. Better Context: Images appear alongside statistical results
  4. No File Management: No need to handle file paths or external viewers
  5. Responsive: Works in any environment where Claude runs

Backward Compatibility

  • Existing scripts: All existing RMCP scripts continue to work unchanged
  • File paths: Still supported for users who want to save plots
  • API: No breaking changes to tool interfaces

This enhancement makes RMCP visualizations much more accessible and user-friendly, providing immediate visual feedback for statistical analyses directly within your Claude conversation.