# ====================================================
# main.py - Diabetes Prediction API (Production)
# ====================================================

from fastapi import FastAPI
from fastapi.middleware.cors import CORSMiddleware
from pydantic import BaseModel
import numpy as np
import joblib
import os

# -----------------------------
# Load Trained Model & Scaler
# -----------------------------
MODEL_FILE = "diabetes_model.pkl"
SCALER_FILE = "scaler.pkl"

if not os.path.exists(MODEL_FILE) or not os.path.exists(SCALER_FILE):
    raise FileNotFoundError("Model or scaler file not found. Make sure both exist in the same directory.")

scaler = joblib.load(SCALER_FILE)
best_model = joblib.load(MODEL_FILE)

# -----------------------------
# Initialize FastAPI app
# -----------------------------
app = FastAPI(
    title="Diabetes Prediction API",
    description="Predicts diabetes based on patient data using trained ML model",
    version="1.0"
)

# -----------------------------
# Enable CORS for all origins
# -----------------------------
app.add_middleware(
    CORSMiddleware,
    allow_origins=["*"],  # For production, replace '*' with allowed domains
    allow_methods=["*"],
    allow_headers=["*"],
)

# -----------------------------
# Define Input Data Model
# -----------------------------
class InputData(BaseModel):
    Age: float
    Sex: float
    BMI: float
    Glucose: float
    BloodPressure: float
    Insulin: float
    Increased_Thirst: float
    Increased_Hunger: float
    Fatigue_Tiredness: float
    Blurred_Vision: float
    Unexplained_Weight_Loss: float

# -----------------------------
# Prediction Endpoint
# -----------------------------
@app.post("/predict")
def predict(data: InputData):
    try:
        # Convert input to numpy array
        features = np.array([[ 
            data.Age,
            data.Sex,
            data.BMI,
            data.Glucose,
            data.BloodPressure,
            data.Insulin,
            data.Increased_Thirst,
            data.Increased_Hunger,
            data.Fatigue_Tiredness,
            data.Blurred_Vision,
            data.Unexplained_Weight_Loss
        ]])

        # Scale features
        features_scaled = scaler.transform(features)

        # Predict
        prediction = best_model.predict(features_scaled)[0]
        result = "Diabetes" if prediction == 1 else "No Diabetes"

        return {"prediction": result}
    
    except Exception as e:
        return {"error": str(e)}

# -----------------------------
# Run the API locally
# -----------------------------
if __name__ == "__main__":
    import uvicorn
    uvicorn.run("main:app", host="0.0.0.0", port=7860, reload=True)