Installation Guide

Step 1 — Extract and Navigate

Extract the downloaded zip file. The main Django project is inside the mumbai_aqi/ subfolder. The Jupyter training notebook is in the root aqi project/ folder.

cd "aqi project/mumbai_aqi"

Step 2 — Create a Virtual Environment

Create and activate a Python virtual environment to keep dependencies isolated.

python -m venv venv

# On Windows
venv\Scripts\activate

# On Mac or Linux
source venv/bin/activate

Step 3 — Install Dependencies

pip install -r requirements.txt

This installs Django, XGBoost, scikit-learn, pandas, numpy, joblib, and all other required packages.

Step 4 — Get a Windy API Key

Visit https://api.windy.com/keys and create a free account. When creating your key, select Point Forecast API. Copy the key — you will need it in the next step.

Step 5 — Configure Environment Variables

cp .env.example .env

Open the .env file in a text editor and fill in the following values:

WINDY_API_KEY=your_windy_api_key_here
SECRET_KEY=your_django_secret_key_here
DEBUG=True
ALLOWED_HOSTS=127.0.0.1,localhost

Step 6 — Train the ML Model

Open mumbai_aqi_model.ipynb in Jupyter Notebook or JupyterLab from the root aqi project/ folder and run all cells. This trains the XGBoost model and saves three files into the models/ directory:

• best_aqi_model.pkl
• scaler.pkl
• feature_columns.pkl

Step 7 — Copy Model Files into Django Project

cp ../models/best_aqi_model.pkl ml_models/
cp ../models/scaler.pkl ml_models/
cp ../models/feature_columns.pkl ml_models/

On Windows, use copy instead of cp, or drag the files manually into the ml_models/ folder inside mumbai_aqi/.

Step 8 — Run Database Migrations

python manage.py migrate

This creates the SQLite database and sets up the StationData and AQIPrediction tables.

Step 9 — Load Station Data (Optional)

python manage.py load_station_data

This populates the StationData table from the included CSV file and enables the 90-day trend charts on the dashboard.

Step 10 — Start the Development Server

python manage.py runserver

Open your browser and go to http://127.0.0.1:8000. The dashboard should load with weather tiles and AQI trend charts.

Step 11 — Run Tests (Optional)

python manage.py test predictor

This runs the unit tests for models, views, and ML utilities.