# app.py
import pandas as pd
from xgboost import XGBRegressor
from sklearn.model_selection import train_test_split
import requests
import os

# Fetch data from Facebook API
def fetch_data_from_api(query, geo_locations):
    url = f"https://graph.facebook.com/v17.0/act_597540533213624/targetingsearch"
    params = {
        "q": query,
        "geo_locations[countries]": geo_locations,
        "access_token": os.getenv('ACCESS_TOKEN')
    }
    response = requests.get(url, params=params)
    if response.status_code == 200:
        return response.json()
    else:
        raise Exception(f"Failed to fetch data from API. Status code: {response.status_code}")

# Generate synthetic metrics
def generate_synthetic_metrics(data):
    IMPRESSION_RATE = 0.10  # 10% of audience sees the ad
    CTR = 0.05  # 5% of impressions result in clicks
    CONVERSION_RATE = 0.02  # 2% of clicks result in conversions
    CPM = 5  # $5 per 1000 impressions
    REVENUE_PER_CONVERSION = 50  # $50 per conversion

    data['impressions'] = data['audience_size_lower_bound'] * IMPRESSION_RATE
    data['clicks'] = data['impressions'] * CTR
    data['conversions'] = data['clicks'] * CONVERSION_RATE
    data['ad_spend'] = (data['impressions'] / 1000) * CPM
    data['revenue'] = data['conversions'] * REVENUE_PER_CONVERSION
    data['roi'] = (data['revenue'] - data['ad_spend']) / data['ad_spend']

    return data

# Train and save the model
def train_and_save_model():
    # Fetch data
    response_data = fetch_data_from_api('Fitness', 'NG')
    data = pd.DataFrame(response_data['data'])

    # Generate synthetic metrics
    data = generate_synthetic_metrics(data)

    # Features and target
    X = data[['ad_spend', 'impressions', 'clicks', 'conversions']]
    y = data['roi']

    # Train the model
    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
    model = XGBRegressor(n_estimators=100, max_depth=3, n_jobs=-1)
    model.fit(X_train, y_train)

    # Save the model
    model.save_model('model.json')
    print("Model saved to 'model.json'.")

if __name__ == '__main__':
    train_and_save_model()