Installation Guide

Clone or Extract the Project

Download the project ZIP from CodeAj Marketplace and extract it, or clone the repository using the command below. Navigate into the project folder after extraction.

git clone <repository_url>
cd "Crop Yield Prediction"

Download the Dataset

The dataset files are hosted externally due to their size. Click the button below to download the dataset from Google Drive. Once downloaded, extract the contents and place all CSV files inside the dataset/ folder in the project root.

Dataset folder should contain:
  dataset/
  ├── pesticides.csv
  ├── rainfall.csv
  ├── temp.csv
  ├── yield_df.csv
  └── yield.csv

Create a Python Virtual Environment

It is strongly recommended to create an isolated virtual environment before installing dependencies. This prevents conflicts with other Python projects on your system.

python -m venv venv

Activate the Virtual Environment

Activate the virtual environment based on your operating system before proceeding to install dependencies.

.\venv\Scripts\activate

source venv/bin/activate

Once activated, you will see (venv) appear at the beginning of your terminal prompt, confirming the environment is active.

Install All Dependencies

Run the following command to install all required Python libraries including Flask, scikit-learn, Keras, TensorFlow, NumPy, Pandas, Matplotlib, and Seaborn in one step.

pip install -r requirements.txt

This step may take 3 to 8 minutes depending on your internet connection as it downloads TensorFlow and other large libraries.

Verify Pre-trained Model Files

The pre-trained machine learning models are included in the models/ directory and are ready to use without any additional training. Verify that the following files are present.

models/
  ├── yield_prediction_model.pkl     (Random Forest — crop yield)
  ├── yield_scaler.pkl               (scaler for yield inputs)
  ├── disease_prediction_model.pkl   (CNN — plant disease detection)
  ├── disease_scaler.pkl             (scaler for disease inputs)
  ├── disease_encoder.pkl            (label encoder for diseases)
  └── weather_model.pkl              (Linear Regression — rainfall)

The fertilizer_recommender.pkl file is optional. If not present, the system automatically falls back to a built-in rule-based recommendation engine. This is expected behavior and does not affect functionality.

Retrain Models (Optional)

If you want to retrain the models from scratch or explore the training process for your project presentation, open the Jupyter Notebook and execute all cells in order.

jupyter notebook training.ipynb

Running the full training notebook will overwrite the existing model files in the models/ directory with newly trained versions. Skip this step if you want to use the pre-trained models.

Run the Flask Application

Start the AgriSmart web server using the command below. The application will launch on your local machine and be accessible through your browser.

python app.py

Once running, open your browser and go to:

http://127.0.0.1:5000/

Navigate the Application

Use the following URLs to access each module of AgriSmart directly from your browser.

API Endpoints (For Developers)

AgriSmart also exposes REST API endpoints that accept POST requests, allowing programmatic access to all four AI modules. Refer to app.py for the exact request payload and response structure for each endpoint.

POST  /api/predict-yield
POST  /api/predict-disease
POST  /api/recommend-fertilizer
POST  /api/predict-weather