import gradio as gr
import pickle
import json
import numpy as np

# Load model and columns
with open("banglore_home_prices_model.pickle", "rb") as f:
    model = pickle.load(f)

with open("columns.json", "r") as f:
    data_columns = json.load(f)["data_columns"]

locations = data_columns[3:]  # Extract location columns (FROM FORTH COLUMN TO END FOUND IN LOCATION)

def predict_price(total_sqft, bath, bhk, location):
    # Prepare the input array
    x = np.zeros(len(data_columns))
    x[0] = total_sqft
    x[1] = bath
    x[2] = bhk
    if location in locations:
        loc_index = data_columns.index(location)
        x[loc_index] = 1
    
    # Make prediction
    return model.predict([x])[0]

# Create the Gradio interface
inputs = [
    gr.Number(minimum=1,label="Total Square Feet"),
    gr.Number(minimum=1,label="Bath"),
    gr.Number(minimum=1,label="BHK [Bedroom, Hall, and Kitchen]"),
    gr.Dropdown(choices=locations, label="Location")
]

outputs = gr.Textbox(label="Predicted Price (Lakh)")

# Add the link under the output prediction
#link = gr.Markdown("For more details, visit [Github](https://github.com/94etienne/AI_PROJECTS_RESEARCH/blob/main/1_banglore_home_price.rar)")


# Footer content
footer = "Etienne NTAMBARA @AI_Engineer"

# Launch the interface
gr.Interface(fn=predict_price, inputs=inputs, outputs=outputs, title="Real Estate Price Prediction", article=footer).launch()