import joblib
import pandas as pd
import xgboost as xgb
from xgboost import XGBRegressor
import numpy as np
from flask import Flask, request, jsonify

# Initialize Flask app with a name
app = Flask(__name__)

# Load the trained churn prediction model
model = joblib.load("superkart_prediction_model_v1_0.joblib")

# Define a route for the home page
@app.get('/')
def home():
    return "Welcome to the Superkart Sales Forecast API"

# Define an endpoint to predict churn for a single customer
@app.route('/v1/product', methods=['POST'])

def predict_forecast():
    """
    This function handles POST requests to the '/v1/product' endpoint.
    It expects a JSON payload containing property details and returns
    the predicted rental price as a JSON response.
    """

    print("🚀 /v1/product route hit")

    # Get JSON data from the request
    superKart_data = request.get_json()

    # Extract relevant customer features from the input data
    sample = {
        'Product_Id': superKart_data['Product_Id'],
        'Product_Weight': superKart_data['Product_Weight'],
        'Product_Sugar_Content': superKart_data['Product_Sugar_Content'],
        'Product_Allocated_Area': superKart_data['Product_Allocated_Area'],
        'Product_Type': superKart_data['Product_Type'],
        'Product_MRP': superKart_data['Product_MRP'],
        'Store_Id': superKart_data['Store_Id'],
        'Store_Establishment_Year': superKart_data['Store_Establishment_Year'],
        'Store_Size': superKart_data['Store_Size'],
        'Store_Location_City_Type': superKart_data['Store_Location_City_Type'],
        'Store_Type': superKart_data['Store_Type']
     }

    # Convert the extracted data into a DataFrame
    input_data = pd.DataFrame([sample])

    predicted_sales_value = model.predict(input_data)[0]  # Avoid .tolist()[0]

    # Clip BEFORE applying exp
    predicted_sales = np.clip(predicted_sales_value, -100, 100)

    # Apply exp safely
    #predicted_sales = np.exp(safe_value)

    # Calculate actual price
    #predicted_sales = np.exp(predicted_sales_value)

    # Convert predicted_price to Python float
    predicted_sales = round(float(predicted_sales), 2)

    # Return the actual price
    return jsonify({'Predicted Sales': predicted_sales})

    # # Map prediction result to a human-readable label
    # prediction_label = "churn" if prediction == 1 else "not churn"

    # # Return the prediction as a JSON response
    # return jsonify({'Prediction': prediction_label})


# Define an endpoint to predict churn for a single customer
#@app.route('/v1/product', methods=['POST'])

# Define an endpoint to predict churn for a batch of customers
@app.route('/v1/superKartBatch', methods=['POST'])

def predict_churn_batch():

    """
    This function handles POST requests to the '/v1/superKartBatch' endpoint.
    It expects a CSV file containing property details for multiple properties
    and returns the predicted rental prices as a dictionary in the JSON response.
    """

    # Get the uploaded CSV file from the request
    file = request.files['file']

    # Read the file into a DataFrame
    input_data = pd.read_csv(file)

    # Make predictions for all properties in the DataFrame (get log_sales)
    predicted_log_sales = model.predict(input_data).tolist()
    print("Raw predictions:", predicted_log_sales)

    safe_log_sales = np.clip(predicted_log_sales, -100, 100)
    predicted_sales = [round(float(np.exp(log_price)), 2) for log_price in safe_log_sales]

    # Calculate actual sales
    predicted_sales = [round(float(np.exp(log_price)), 2) for log_price in predicted_log_sales]

    # Create a dictionary of predictions with property IDs as keys
    if 'ProductId' in input_data.columns:
        product_id_list = input_data['ProductId'].tolist()
    elif 'Product_Id' in input_data.columns:
        product_id_list = input_data['Product_Id'].tolist()
    else:
        product_id_list = [f"Product_{i}" for i in range(len(predicted_sales))]

    output_dict = dict(zip(product_id_list, predicted_sales))

    #raw_pred = model.predict(X)
    #print("Raw:", raw_pred)

    for rule in app.url_map.iter_rules():
      print(f"🔗 Route: {rule} → Methods: {rule.methods}")

    # Return the predictions dictionary as a JSON response
    return output_dict