app.py

import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression, Ridge, Lasso
from sklearn.metrics import r2_score
import gradio as gr


def regularization(upload, l2=None, l1=None):
    
    if upload is not None:
        
        # Daten einlesen und aufteilen
        df = pd.read_csv(upload)
        X = df.drop(['target'], axis=1)
        Y = df[['target']]

        # Training LINEARE REGRESSION
        lr = LinearRegression()
        X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.4, random_state=1234)
        lr.fit(X_train, Y_train)

        # Vorhersagen TRAINING & TEST
        Y_predict_train = lr.predict(X_train)
        Y_predict_test = lr.predict(X_test)

        # Kennzahlen TRAINING & TEST
        r2_train = r2_score(Y_train, Y_predict_train)
        r2_test = r2_score(Y_test, Y_predict_test)
        result_linear = f"LINEARE REGRESSION\nR2 Training: {r2_train:.2f}\nR2 Test: {r2_test:.2f}\n\n"

        result_ridge = ""
        result_lasso = ""

        if l2:
            # RIDGE Regression
            ridge = Ridge(alpha=float(l2))
            ridge.fit(X_train, Y_train)

            # Vorhersagen TRAINING & TEST
            Y_predict_train_l2 = ridge.predict(X_train)
            Y_predict_test_l2 = ridge.predict(X_test)
            
            # Kennzahlen TRAINING & TEST
            r2_train_l2 = r2_score(Y_train, Y_predict_train_l2)
            r2_test_l2 = r2_score(Y_test, Y_predict_test_l2)
            result_ridge = f"RIDGE REGRESSION\nR2 Training: {r2_train_l2:.2f}\nR2 Test: {r2_test_l2:.2f}\n\n"
        
        if l1:
            # LASSO Regression
            lasso = Lasso(alpha=float(l1))  # Hier die fehlende Klammer hinzugefügt
            lasso.fit(X_train, Y_train)

            # Vorhersagen TRAINING & TEST
            Y_predict_train_l1 = lasso.predict(X_train)
            Y_predict_test_l1 = lasso.predict(X_test)
            
            # Kennzahlen TRAINING & TEST
            r2_train_l1 = r2_score(Y_train, Y_predict_train_l1)
            r2_test_l1 = r2_score(Y_test, Y_predict_test_l1)
            result_lasso = f"LASSO REGRESSION\nR2 Training: {r2_train_l1:.2f}\nR2 Test: {r2_test_l1:.2f}"
        
        result = result_linear + result_ridge + result_lasso
        return result

    else:
        return "Bitte eine Datei hochladen!"



io1 = gr.Interface(
    title = 'L1 & L2 Regularization',
    description = '''Demo APP''',
    fn=regularization, 
    inputs=[gr.File(label="Upload Data"), 
            gr.Text(label="L2 Regularization (Ridge)", placeholder=" L2 eingeben"),
            gr.Text(label="L1 Regularization (Lasso)", placeholder="L1 eingeben")],
    outputs=gr.Textbox(lines=50, label='R-Squared (Coefficient of Determination)'),
    api_name="Regularization"
)

io1.launch()