How mihirj2002/heartdisease_machineleaning_model works
This repository is a heart-disease classification workflow: it loads patient data, prepares features, trains a classifier, and reports evaluation metrics. The sampled code shows a Random Forest implementation in src/model_building.py; although the README describes exploratory analysis and AutoGluon, those implementations are not present in the sampled source, so they are not represented as runtime components. The loader reads from a remote CSV URL, while a CSV also exists in the repository; whether the local file is used is unclear.
Main entry points: main.py Pipeline entry · data_preprocessing.py Preprocess features · data_ingestion.py Data ingestion · model_building.py Random Forest
Components 11
Workflow entry
- Pipeline entry
main.py
Data preparation
- Data ingestion
src/data_ingestion.py - Preprocess features
src/data_preprocessing.py - Train-test split
src/data_preprocessing.py - Feature scaling
src/data_preprocessing.py - Class balancing
src/data_preprocessing.py
Modeling
- Random Forest
src/model_building.py - Model evaluation
src/model_building.py
Other components
- Workflow invoker
- Remote CSV
- Heart dataset
data/heart_disease_risk_2026.csv
Connections 12
- Workflow invoker → Pipeline entry: invokes
- Pipeline entry → Data ingestion: calls
- Remote CSV → Data ingestion: provides data
- Data ingestion → Pipeline entry: returns data
- Pipeline entry → Preprocess features: calls
- Preprocess features → Train-test split: separates features
- Train-test split → Feature scaling: scales sets
- Feature scaling → Class balancing: balances training
- Preprocess features → Pipeline entry: returns sets
- Pipeline entry → Random Forest: calls
- Random Forest → Model evaluation: predicts and scores
- Model evaluation → Pipeline entry: prints metrics