Cross Val and Bayesian hyperparameter tuning for Random Forest

crossval_bayesian_rf.py

from bayes_opt import BayesianOptimization

def RF_opt(n_estimators, max_depth): 
    
  global rskf 
  reg = RandomForestClassifier(verbose = 0,
                          n_estimators = int(n_estimators),
                          #min_samples_split = int(min_samples_split),
                          #min_samples_leaf = int(min_samples_leaf),
                          max_depth = int(max_depth),
                          warm_start = True, 
                          #max_features = int(max_features),
                          random_state = rs)
  
  reg.fit(X1,Y1)
  scores = cross_val_score(reg, X1, Y1, scoring = mcc, cv = rskf)
  return np.mean(scores)

pbounds = {"n_estimators": (40, 500),
          "max_depth": (2,14),
          #"min_samples_split": (10, 20),
          #"min_samples_leaf":(10, 20),
          #"max_features": None,
           }


optimizer = BayesianOptimization(
    f = RF_opt,
    pbounds = pbounds,
    verbose = 2,
    
)

optimizer.maximize(init_points = 2, n_iter = 20)

print(optimizer.max)