import gradio as gr
import pandas as pd
import lightgbm as lgb
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import LabelEncoder

# Load dataset
url = "https://raw.githubusercontent.com/Pushpinder-Singh06/CSV-Files/refs/heads/main/crop_cleaned%20data.csv "
data = pd.read_csv(url)

# Define the crops in Punjabi
crops = [
    'rice',
    'maize',
    'chickpea',
    'kidneybeans',
    'pigeonpeas',
    'mothbeans',
    'mungbean',
    'blackgram',
    'lentil',
    'pomegranate',
    'banana',
    'mango',
    'watermelon',
    'muskmelon',
    'orange',
    'papaya',
    'coconut',
    'cotton',
    'jute',
    'coffee'
]

# Fit the label encoder
le = LabelEncoder()
le.fit(data['label'])  # ✅ Fit first
data['crop_encoded'] = le.transform(data['label'])  # ✅ Then transform

# Prepare for reverse prediction
reverse_X = data[['crop_encoded']]
y_cols = ['N', 'P', 'K', 'temperature', 'humidity', 'ph', 'rainfall']


# Train reverse models
reverse_models = {}
for col in y_cols:
    y = data[col]
    X_train, X_test, y_train, y_test = train_test_split(reverse_X, y, test_size=0.2, random_state=42)
    model_r = lgb.LGBMRegressor()
    model_r.fit(X_train, y_train)
    reverse_models[col] = model_r

# Mapping from crop name to encoded value
label_to_encoded = {label: le.transform([label])[0] for label in le.classes_}

# Reverse prediction function
def predict_crop_parameters(crop_name):
    crop_name = crop_name.strip()
    if crop_name not in label_to_encoded:
        return f"❌ No data found for '<b>{crop_name}</b>'."

    encoded_value = label_to_encoded[crop_name]
    input_data = [[encoded_value]]

    predictions = {}
    for param, model_r in reverse_models.items():
        predicted_value = model_r.predict(input_data)[0]
        predictions[param] = round(predicted_value, 2)

    formatted_output = ""
    for k, v in predictions.items():
        formatted_output += f"<b>{k}</b>: {v}<br>"

    return formatted_output


    
with gr.Blocks() as demo:
    gr.Markdown("# 🌾 **AgroVision: Crop-Based Parameters**")
    crop_input = gr.Dropdown(choices=crops, label="🌿 Select Crop Name")
    result_output = gr.Markdown(label="🧪 Recommended Parameters")
    run_btn = gr.Button("➡️ Predict Parameters")
    run_btn.click(fn=predict_crop_parameters, inputs=[crop_input], outputs=[result_output])


demo.launch()