[MRG] Numpy error with transformers.py

[Tekom]

Most recent numpy version works with np.int64 or np.int32 insted of just np.int, thats was generating an error when using class spaces.

#Test
from sklearn.ensemble import GradientBoostingClassifier
from skopt import BayesSearchCV
from skopt.space import Real, Categorical, Integer
from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split

Cargar el dataset de iris

iris = load_iris()

Dividir los datos en entrenamiento y prueba

X_train, X_test, y_train, y_test = train_test_split(iris.data, iris.target, test_size=0.3, random_state=42)

Definir el modelo de árbol de decisiones

clf = GradientBoostingClassifier()

param_space = {
'learning_rate': Real(1e-5, 1, prior = 'log-uniform'),
'n_estimators': Integer(20, 1_500),
'subsample': Real(0.05, 1),
'max_depth': Integer(1, 10),
}

Realizar la búsqueda bayesiana

bayes_dt = BayesSearchCV(clf, param_space, n_iter=50, cv=5, n_jobs=-1)

Ajustar el modelo

bayes_dt.fit(X_train, y_train)

Imprimir los mejores hiperparámetros y la puntuación del modelo

print('Mejores hiperparámetros:', bayes_dt.best_params_)
print('Puntuación del modelo:', bayes_dt.best_score_)

#End of test

If you keep using np.int insted of np.int64 or np.int32 thats will show an error saying that numpy has no method called int.