How manivasagam-rajakannan/global-development-clustering works
This is a Streamlit dashboard for exploring country-development clusters. The principal workflow is CSV upload, country-level cleaning and aggregation, scaling, selectable clustering, metric evaluation, then visual and tabular exploration or result download. The app also supports country lookup and predictions for hypothetical countries. The README describes a companion notebook analysis, but its implementation was not sampled, so its internal stages and connections are not inferred here; the graph’s source-backed edges come from app.py.
Main entry points: app.py Streamlit Dashboard
Components 14
Input Preparation
- Clean and Aggregate
app.py - Feature Scaling
app.py
Clustering Analysis
- PCA Projection
app.py - Clustering Models
app.py - Cluster Metrics
app.py
Results Exploration
- PCA Scatter Plot
app.py - World Cluster Map
app.py - Cluster Profiles
app.py
User Tools
- Streamlit Dashboard
app.py - Country Lookup
app.py - New Country Prediction
app.py - CSV Results Download
app.py
Other components
- Dashboard User
- Development CSV
Connections 27
- Dashboard User → Streamlit Dashboard: uses
app.py - Development CSV → Streamlit Dashboard: uploads
app.py - Streamlit Dashboard → Clean and Aggregate: cleans and aggregates
app.py - Streamlit Dashboard → Feature Scaling: scales features
app.py - Streamlit Dashboard → PCA Projection: projects features
app.py - Streamlit Dashboard → Clustering Models: fits and assigns
app.py - Streamlit Dashboard → Cluster Metrics: evaluates labels
app.py - Streamlit Dashboard → PCA Scatter Plot: renders projection
app.py - Streamlit Dashboard → World Cluster Map: renders assignments
app.py - Streamlit Dashboard → Cluster Profiles: computes and displays
app.py - Streamlit Dashboard → Country Lookup: looks up cluster
app.py - Streamlit Dashboard → New Country Prediction: handles prediction
app.py - Streamlit Dashboard → CSV Results Download: offers clustered CSV
app.py - Clean and Aggregate → Feature Scaling: provides features
app.py - Feature Scaling → PCA Projection: transforms features
app.py - Feature Scaling → Clustering Models: transforms features
app.py - Clustering Models → Cluster Metrics: provides labels
app.py - Clustering Models → PCA Scatter Plot: provides labels
app.py - PCA Projection → PCA Scatter Plot: provides coordinates
app.py - Clustering Models → World Cluster Map: provides assignments
app.py - Clean and Aggregate → Cluster Profiles: provides features
app.py - Clustering Models → Cluster Profiles: provides labels
app.py - Clustering Models → Country Lookup: provides assignments
app.py - Feature Scaling → New Country Prediction: transforms input
app.py - Clustering Models → New Country Prediction: predicts or finds neighbor
app.py - Clean and Aggregate → CSV Results Download: provides features
app.py - Clustering Models → CSV Results Download: provides labels
app.py