import gradio as gr
import pandas as pd
from sklearn.linear_model import LogisticRegression

def create_dummy_data():
    import numpy as np
    np.random.seed(42)
    n = 1000
    data = pd.DataFrame({
        'V1': np.random.randn(n),
        'V2': np.random.randn(n),
        'V3': np.random.randn(n),
        'V4': np.random.randn(n),
        'V5': np.random.randn(n),
        'Amount': np.random.rand(n) * 100,
        'Class': [0]*950 + [1]*50
    })
    return data

def train_model():
    df = create_dummy_data()
    X = df.drop('Class', axis=1)
    y = df['Class']
    model = LogisticRegression(max_iter=1000)
    model.fit(X, y)
    return model

model = train_model()

def predict_fraud(file):
    df = pd.read_csv(file.name)
    X = df.drop(columns=['Class'], errors='ignore')
    preds = model.predict(X)
    df['Prediction'] = preds
    df['Prediction'] = df['Prediction'].map({0: "Legitimate", 1: "Fraudulent"})
    return df

iface = gr.Interface(
    fn=predict_fraud,
    inputs=gr.File(label="Upload CSV file (with columns: V1, V2, V3, V4, V5, Amount)"),
    outputs=gr.Dataframe(label="Predictions"),
    title="Credit Card Fraud Detection Demo",
    description="Upload a CSV file with features V1-V5 and Amount to see if transactions are fraudulent or legitimate."
)

iface.launch()