machine_error.py

import time
import numpy as np
import logging
import time

start = time.time()

from sklearn.linear_model import SGDClassifier
from sklearn.datasets import make_classification
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score, log_loss

# ---------------- LOGGER ----------------
logging.basicConfig(
    level=logging.INFO,
    format='[%(levelname)s] %(message)s'
)

logger = logging.getLogger(__name__)

logger.info("Corriendo machine_MAL.py...")

logger.info("Iniciando entrenamiento...")

# ---------------- DATA ----------------
X, y = make_classification(
    n_samples=200000,
    n_features=20,
    random_state=42
)

X_train, X_val, y_train, y_val = train_test_split(
    X, y, test_size=0.2, random_state=42
)

# ---------------- MODEL ----------------
model = SGDClassifier(loss="log_loss", random_state=42)
epochs = 60

for epoch in range(epochs):
    logger.info(f"Época {epoch+1}/{epochs}")

    # chequeo de datos antes de entrenar
    if np.isnan(X_train).any() or np.isinf(X_train).any():
        logger.error("Dataset corrupto detectado (NaN/Inf). Deteniendo entrenamiento.")
        break

    try:
        model.partial_fit(X_train, y_train, classes=np.unique(y))

        train_acc = accuracy_score(y_train, model.predict(X_train))
        val_acc = accuracy_score(y_val, model.predict(X_val))

        train_loss = log_loss(y_train, model.predict_proba(X_train))
        val_loss = log_loss(y_val, model.predict_proba(X_val))

        logger.info(
            f"Train acc: {train_acc:.4f} | loss: {train_loss:.4f} || "
            f"Val acc: {val_acc:.4f} | loss: {val_loss:.4f}"
        )

    except Exception as e:
        logger.error(f"Error crítico en época {epoch+1}: {e}")
        break  # aquí se detiene el entrenamiento

    # simulación de fallo
    if epoch == 30:
        logger.error("Corrupción de datos introducida")
        X_train[0:1000] = np.nan

    time.sleep(0.5)

end = time.time()

logger.info(f"Tiempo total: {round(end - start, 2)} segundos")
logger.info("Proceso terminado")