Diabetes Risk Prediction

The Diabetes Risk Prediction system is a tool designed to assess an individual's likelihood of developing diabetes based on various health metrics and lifestyle factors. By analyzing data such as age, blood glucose levels, BMI, and other factors, the system provides personalized risk assessments and recommendations for prevention.

Overview

This project is organized into several parts:

• Data Processing & Visualization
  The data is sourced from data/diabetes.csv and preprocessed in src/train_model.py. A bar chart visualizing the diabetes outcome distribution is generated and saved to static/chart.png, then rendered in the templates/index.html file.

• Machine Learning Model
  A RandomForestClassifier is trained using scaled features, and both the model and scaler are saved as src/model/rf_model.pkl and src/model/scaler.pkl respectively.

• API Service
  The FastAPI based API is implemented in api/main.py. This renders the result visualization and serves static files. To run the API, simply launch the provided FastAPI server.

• Command-Line Prediction
  The src/predict.py script allows for console-based predictions where a user inputs health parameters and the system outputs whether the patient is predicted as diabetic or not.

Installation

1. Clone the Repository

   git clone <repository-url>
   cd Diabetes-risk-prediction

2. Set Up a Virtual Environment

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install Dependencies

   Make sure you have the required packages installed. For example, install dependencies such as pandas, scikit-learn, joblib, fastapi, uvicorn, and matplotlib:

   pip install pandas scikit-learn joblib fastapi uvicorn matplotlib

Workflow

1. Data Preparation & Model Training

   Run the training script to preprocess data, train the model, and generate the visualization:

   python src/train_model.py

   This script reads data from data/diabetes.csv, scales the features, splits the dataset, trains a Random Forest model, prints the model accuracy, and saves both the model and scaler in the src/model directory. It also generates and saves a chart in static/chart.png.

2. Running the API

   Launch the FastAPI application:

   uvicorn api.main:app --reload

   Then open your browser at http://127.0.0.1:8000 to view the visualization rendered by templates/index.html.

3. Making Predictions from the Command Line

   Run the prediction script to provide patient data and receive a risk prediction:

   python src/predict.py

   Follow the prompts to input patient data. The script will output whether the patient is predicted as diabetic or not.

Additional Notes

• Data file: data/diabetes.csv
• Visualization: Saved as static/chart.png and rendered in templates/index.html
• API Entry Point: api/main.py
• Model Training and Prediction: src/train_model.py and src/predict.py

Happy Coding!