import gradio as gr
import pickle
import numpy as np
import os

# Load the RandomForest model
working_dir = os.path.dirname(os.path.abspath(__file__))
model = pickle.load(open(f'{working_dir}/RF_Crop.sav', 'rb'))

# Overview section content
overview_text = """
### Welcome to the Crop Recommendation App!

This application assists farmers in selecting the optimal crop to cultivate, considering soil composition
and environmental conditions. By providing information such as nitrogen, phosphorus, and
potassium levels, as well as temperature, humidity, pH, and rainfall, users receive tailored
recommendations for the most suitable crop out of a selection of 22 options.

### How to Use the App
1. Navigate to the "Crop Recommendation" section.
2. Enter the values for the soil and environmental factors in the input fields.
3. Click the "Predict" button to get the crop recommendation.

### About the Model
The recommendation is made using a Random Forest model trained on agricultural data.
This model considers various factors to predict the best crop for your field.
The model has been developed by analyzing many models like SVM, Random Forest,
Decision Tree, Logistic Regression, Gaussian Naive Bayes. Random Forest has been selected based on
the Cross Validation Accuracy & Test Accuracy.

### Benefits of Using Crop Recommendation
- **Increased Yield**: By planting the most suitable crop, you can maximize your harvest.
- **Cost Efficiency**: Avoid wasting resources on crops that are not suited to your soil and climate.
- **Sustainable Farming**: Promote better land use and reduce environmental impact.

### Contact Us
If you have any questions or feedback about the project, feel free to reach out:
- **Email**: kanchanrai2307@gmail.com
- **Github**: [kanchanrai7](https://github.com/kanchanrai7)
"""

# Define the prediction function
def predict_crop(N, P, K, temperature, humidity, pH, rainfall):
    user_input = np.array([[N, P, K, temperature, humidity, pH, rainfall]])
    if np.all(user_input == 0):
        return "Please enter valid values."
    else:
        prediction = model.predict(user_input)
        crop = prediction[0]
        return f"Hey, you should grow **{crop}** based on your soil and environmental factors."

# Gradio UI components
def main_interface():
    with gr.Blocks() as demo:
        with gr.Tab("Overview"):
            gr.Markdown(overview_text)
            gr.Image("Images/image1.jpg", label="Healthy Crops")
            gr.Image("Images/mod_comparison.png", label="Model Comparison")
        
        with gr.Tab("Get Recommendation"):
            gr.Markdown("Enter the details about your soil and environmental factors to get a crop recommendation.")
            gr.Markdown("**Example Values:** [104, 18, 30, 23.6, 60.3, 6.7, 140.91] or [60, 18, 30, 23.6, 60.3, 8, 40.91]")

            N = gr.Number(label="Nitrogen (N)", value=0, precision=0)
            P = gr.Number(label="Phosphorus (P)", value=0, precision=0)
            K = gr.Number(label="Potassium (K)", value=0, precision=0)
            temperature = gr.Number(label="Temperature (°C)", value=0.0)
            humidity = gr.Number(label="Humidity (%)", value=0.0)
            pH = gr.Number(label="pH", value=0.0)
            rainfall = gr.Number(label="Rainfall (mm)", value=0.0)

            output = gr.Textbox(label="Recommendation", interactive=False)

            gr.Button("Predict").click(
                predict_crop, 
                inputs=[N, P, K, temperature, humidity, pH, rainfall], 
                outputs=output
            )

    return demo

# Run the Gradio app
if __name__ == "__main__":
    app = main_interface()
    app.launch()