import joblib

from fastapi import FastAPI, Request, Form

from fastapi.responses import HTMLResponse

from fastapi.staticfiles import StaticFiles

from fastapi.templating import Jinja2Templates

from pydantic import BaseModel

import uvicorn

# Load the saved model

model = joblib.load('cancer_model.pkl')

import sklearn

print(sklearn.__version__)

# Define the input data schema

class PredictionInput(BaseModel):

    gender: str = Form()

    age: int = Form()

    smoking: str = Form()

    yellowFingers: str = Form()

    anxiety: str = Form()

    peerPressure: str = Form()

    chronicDisease: str = Form()

    fatigue: str = Form()

    allergy: str = Form()

    wheezing: str = Form()

    alcoholConsuming: str = Form()

    coughing: str = Form()

    shortnessOfBreath: str = Form()

    swallowingDifficulty: str = Form()

    chestPain: str = Form()

# Create the FastAPI application

app = FastAPI()

templates = Jinja2Templates(directory="static")

# Mount the static files directory to serve the HTML and CSS files

app.mount("/static", StaticFiles(directory="static"), name="static")

# Define the prediction route

@app.get("/", response_class=HTMLResponse)

async def home(request: Request):

    return templates.TemplateResponse("/index.html", {"request": request})

@app.post("/predict")

async def predict_lung_cancer(

    request: Request,

    gender: str = Form(),

    age: int = Form(),

    smoking: str = Form(),

    yellowFingers: str = Form(),

    anxiety: str = Form(),

    peerPressure: str = Form(),

    chronicDisease: str = Form(),

    fatigue: str = Form(),

    allergy: str = Form(),

    wheezing: str = Form(),

    alcoholConsuming: str = Form(),

    coughing: str = Form(),

    shortnessOfBreath: str = Form(),

    swallowingDifficulty: str = Form(),

    chestPain: str = Form()

):

    # Convert the input data to a dictionary

    input_data = {

        # "gender":gender,

        # "age":float(age),

        # "smoking":float(smoking),

        "yellowFingers":float(yellowFingers),

        "anxiety":float(anxiety),

        "peerPressure":float(peerPressure),

        "chronicDisease":float(chronicDisease),

        "fatigue":float(fatigue),

        "allergy":float(allergy),

        "wheezing":float(wheezing),

        "alcoholConsuming":float(alcoholConsuming),

        "coughing":float(coughing),

        "shortnessOfBreath":float(shortnessOfBreath),

        "swallowingDifficulty":float(swallowingDifficulty),

        "chestPain":float(chestPain)

    }

    print(type(input_data))

    # Prepare the input features for prediction

    input_features = [input_data[col] for col in input_data]

    # Make the prediction

    prediction = model.predict([input_features])[0]

    # Return the prediction as a JSON response

    return {"prediction": int(prediction)}

if __name__ == "__main__":

    uvicorn.run(app, host="0.0.0.0", port=8000)