app.py

import pickle
import pandas as pd
from sklearn.linear_model import LogisticRegression
from sklearn.metrics import classification_report, accuracy_score
from sklearn.model_selection import train_test_split
from fastapi import FastAPI, UploadFile, File, HTTPException
from pydantic import BaseModel
import io

app = FastAPI()
data = None

# Function to train the model
def train_aut(data):
    data['Downtime_Flag'] = data['Downtime_Flag'].map({'Yes': 1, 'No': 0})
    X = data[['Temperature', 'Run_Time']]
    y = data['Downtime_Flag']
    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
    model = LogisticRegression()
    model.fit(X_train, y_train)
    with open('model.pkl', 'wb') as file:
        pickle.dump(model, file)
    y_pred = model.predict(X_test)
    accuracy = accuracy_score(y_test, y_pred)
    f1 = classification_report(y_test, y_pred, output_dict=True)['1']['f1-score']
    return accuracy, f1

# Function to make predictions
def predict_aut(temp, run_time):
    try:
        with open('model.pkl', 'rb') as file:
            model = pickle.load(file)
        input_data = [[temp, run_time]]
        y_pred = model.predict(input_data)
        return 'Yes' if y_pred[0] == 1 else 'No'
    except FileNotFoundError:
        raise HTTPException(status_code=400, detail="Model not trained. Please upload data and train the model first.")

# Pydantic model for prediction input
class PredictionInput(BaseModel):
    Temperature: float
    Run_Time: float

@app.post("/upload")
async def upload(file: UploadFile = File(...)):
    try:
        global data
        contents = await file.read()
        data = pd.read_csv(io.StringIO(contents.decode("utf-8")))
        return {"message": "File uploaded successfully."}
    except Exception as e:
        raise HTTPException(status_code=400, detail=f"Error reading file: {str(e)}")

@app.post("/train")
def train():
    global data
    if data is None:
        raise HTTPException(status_code=400, detail="No data uploaded. Please upload a dataset first.")
    try:
        accuracy, f1 = train_aut(data)
        # return {"message": "Model trained successfully.", "accuracy": accuracy, "f1_score": f1}
        return {"message": "Please Contact the owner to switch this space on."}
    except Exception as e:
        raise HTTPException(status_code=500, detail=f"Error during training: {str(e)}")

@app.post("/predict")
def predict(input_data: PredictionInput):
    try:
        result = predict_aut(input_data.Temperature, input_data.Run_Time)
        # return {"Downtime": result}
        return {"message": "Please Contact the owner to switch this space on."}
    except Exception as e:
        raise HTTPException(status_code=500, detail=f"Error during prediction: {str(e)}")

if __name__ == "__main__":
    import uvicorn
    uvicorn.run(app, host="0.0.0.0", port=8000)