import sys

from whyteboard import tools



sys.modules['tools'] = tools

import numpy as np



from sklearn.ensemble import GradientBoostingClassifier

from sklearn.model_selection import GridSearchCV

from sklearn.datasets import load_iris



def feature_transformation(gscv, x):

  ret = [gscv.best_estimator_.estimators_[i, 0].tree_.apply(x.astype(np.float32)) for i in

     range(gscv.best_estimator_.n_estimators)]

  return np.array(ret)



def ft_gbdt():

  data = load_iris()



  x = data['data']

  y = data['target']

  x_train = x[0:130]

  x_test = x[130:150]

  y_train = y[0:130]

  y_test = y[130:150]



  print("x:{}".format(np.array(x).shape))

  print("y:{}".format(np.array(y).shape))



  # tune hyper parameters first

    model = GradientBoostingClassifier(n_estimators=1000)

  parameters = {'learning_rate' : [0.1],

                  'max_depth': [4, 6],

                  'min_samples_leaf': [3, 5],

                  'max_features': [1.0]}



  gscv = GridSearchCV(model, parameters, verbose=10, n_jobs=-1, cv=4)

  gscv.fit(x_train, y_train)

  print("score:{}".format(gscv.score(x_test, y_test)))



  # ------------------------------------------------

  ret = feature_transformation(gscv, x_test)



  print("ret:{}".format(ret))