GeoGrad

An interactive geospatial decision-support system for U.S. Master's program selection. GeoGrad integrates five public datasets covering 2,095 four-year institutions across all 50 U.S. states and Washington D.C., providing coordinated multi-view visualization to help prospective graduate students make informed decisions.

Features

• National Choropleth Map with four switchable state-level metrics (cost of living, safety, climate, university count) and a toggleable university overlay showing all 2,095 institutions as clustered markers
• Interactive Filtering via sidebar range sliders (7 metrics) and parallel coordinates axis brushing, with cross-view coordination that propagates filters to all views simultaneously
• Filter-Aware University Overlay on the national map that dynamically excludes non-matching institutions from cluster groups, so cluster counts reflect only filtered results
• State Drill-Down with metric-aware marker clustering at actual GPS coordinates and five switchable university-level metrics
• Scatter Plot with eight selectable axes for exploring bivariate correlations across all institutions
• Parallel Coordinates with seven axes, draggable axis reordering, and axis brushing for multi-dimensional filtering
• Radar Chart comparing selected universities across seven normalized dimensions (normalized against the full 2,095-institution dataset)
• Multi-State Comparison Panel with ten metrics and direction-aware best-value highlighting
• University Comparison Panel with nine metrics, best-value highlighting, and direct website links
• Distance Measurement Tool using Haversine formula for straight-line distances between a reference pin and compared universities
• University Search Bar with autocomplete across all 2,095 institutions

Tech Stack

• React.js v18 with React Router and React Context API for coordinated state management
• Leaflet.js with Leaflet.markercluster for interactive maps, GeoJSON overlays, and marker clustering
• D3.js v7 for color scales, scatter plot, parallel coordinates with axis brushing, radar chart, and enrollment-based marker sizing
• CARTO Positron light tiles as the base map

Data Sources

Source	Level	Description
IPEDS 2023/24	Institution	Tuition, enrollment, admission/graduation/retention rates, GPS coordinates, financial aid, faculty data
Scimago IR	Institution	Overall and research rankings (US and global), fuzzy-matched to IPEDS at 82% threshold
Cost of Living Index	State	Total annual cost, housing cost, median income
FBI UCR	State	Hate-crime incidents and victims (safety proxy)
NOAA Climate	State	Average high/low temperature, precipitation

Pre-generated data files are included in src/data/:

• university_data.json (2,095 records, 1.8 MB)
• state_data.json (51 entries, 32 KB)
• us-states.json (GeoJSON boundaries, 96 KB)

Getting Started

Prerequisites

• Node.js v18 or later
• npm

Installation

cd GeoGrad
npm install

Running

npm start

Opens the application at http://localhost:3000.

Project Structure

GeoGrad/
  src/
    components/
      NationalMap.js        # Choropleth + university overlay + filter-aware clustering
      StateMap.js            # State drill-down with metric-aware marker clustering
      ScatterPlot.js         # Bivariate scatter plot with selectable axes
      ParallelCoordinates.js # 7-axis parallel coordinates with brushing
      RadarChart.js          # Normalized radar chart for university comparison
      FilterPanel.js         # Sidebar range sliders (7 metrics)
      ComparisonPanel.js     # Multi-state comparison panel
      UniversityComparisonPanel.js  # University-level comparison
      UniversityDetailPanel.js      # Individual university detail view
      SearchBar.js           # Autocomplete university search
      Navbar.js              # Navigation bar
    context/
      AppContext.js           # Shared state (13 variables) via React Context
    pages/
      MapPage.js             # Main visualization page (map + sidebar)
      HomePage.js            # Landing page
      AboutPage.js           # About the project
      DataSourcesPage.js     # Data source descriptions
      HowItWorksPage.js      # Usage guide
    data/
      university_data.json   # 2,095 institution records
      state_data.json        # 51 state-level entries
      us-states.json         # GeoJSON state boundaries
    utils/
      distance.js            # Haversine distance calculation
    App.js                   # Root component with routing
    App.css                  # Global styles

Data Pipeline

The data pipeline (build_data.py, ~540 lines) is available in the project root. It:

1. Extracts institution data from the IPEDS 2023/24 Access database (587 MB) using mdb-export
2. Cleans and normalizes ~40 fields per institution
3. Fuzzy-matches institution names to Scimago rankings at 82% threshold
4. Aggregates institution-level metrics to state-level averages
5. Outputs state_data.json and university_data.json

The pipeline requires Python 3 with Pandas and mdb-tools. Since pre-generated JSON files are included, the pipeline does not need to be re-run to use the application.

Author

Chinmay Mhatre - University of Arizona (chinmaymhatre@arizona.edu)

Developed as part of CSC 544: Data Visualization, Spring 2026.