Skip to content

README.md

Latest commit

 

History

History
78 lines (57 loc) · 3.85 KB

README.md

File metadata and controls

78 lines (57 loc) · 3.85 KB

Interactive Decision Map (IDM-style)

Note: All code and documentation in this project have been generated by GitHub Copilot. Human validation has been limited to testing the application functionality.

This viewer lets you explore multi-objective datasets with slider-based brushing across all metrics, interactive visualizations of metric relationships, and detailed inspection of individual solutions. Click a point to highlight it across all views.

Features

  • Tabbed Visualization Layout: Three complementary views to analyze trade-offs
    • Scatter Plot: 2D metric space with Pareto front highlighting
    • Parallel Coordinates: All metrics for brushed points with constraint ranges
    • Radar Plot: Multi-objective profiles as polar dimensions
  • Slider-based Brushing: Filter solutions across all metrics with responsive updates
  • Scaled/Unscaled Metrics: Toggle between normalized [0,1] and raw values
  • Sample Data Loader: Quick-start with built-in test dataset
  • Point Inspection: Click any point to see its parameters and metrics

Gallery

Scatter Plot - Initial View

Scatter plot with sample data loaded

Scatter Plot - With Brushing

Scatter plot with slider brushing applied

Parallel Coordinates - Brushed Data

Parallel coordinates showing all metrics for brushed subset

Radar Plot - Normalized Metrics

Radar plot showing metric profiles in normalized scale

Radar Plot - Raw Values

Radar plot showing metric profiles in raw scale

How to interpret each plot

Scatter Plot

  • X and Y axes: your selected metric_* columns (normalized to [0,1], lower is better).
  • Grey points: all brushed points (i.e., those within the slider ranges across all metrics).
  • Red points: brushed points that are also on the global Pareto front in the full metric space.
  • Blue line: local 2D Pareto front within the brushed subset for the selected (X,Y) metrics.
  • Blue dot: the clicked/selected point.

Parallel Coordinates

  • One vertical axis per metric_*; values are normalized to [0,1] (lower is better).
  • Grey lines: all brushed points.
  • Red lines (when no point is selected): lines that are globally Pareto-optimal.
  • Axis shading (constraint range) reflects your slider brush per metric.

Radar Plot

  • One polar axis per metric_*; values normalized to [0,1] (or raw, see toggle).
  • Semi-transparent traces: all brushed points, colored by the selected dimension.
  • Solid blue trace: the currently selected point (highlighted for clarity).
  • Useful for understanding multi-objective trade-offs and comparing metric profiles across solutions.

Inspection Window (left)

  • Shows param_* (parameters) and metric_* (metrics) for the selected point.
  • Color dimension selector to highlight patterns in the data.

Slider-based Brushing (Recommended)

Use the sliders (left column) to constrain every metric's range. The constraints are applied to all points before plotting, so all views show only the brushed subset. This gives you reliable, responsive filtering without external events.

Metric Scales

Toggle between Normalized [0,1] (for comparing metrics on equal footing) and Raw values (to see original units). Sliders, axes, and all visualizations automatically adjust.

Quick Start

Load sample data by clicking the "Load Sample" button next to the CSV upload box. No file needed—instant exploration!

Or upload your own CSV with columns formatted as:

  • param_*: parameter/design variable columns
  • metric_*: objective/performance metric columns

Run

python idm_viewer/app.py

App runs at: http://0.0.0.0:8050