LL = make_pipeline(RobustScaler(), LassoLars(alpha=0.0001))
LL.data = 0
HR = HuberRegressor(epsilon=1., max_iter=300)
HR.data = 0
GBoost = GradientBoostingRegressor(n_estimators=3000,
learning_rate=0.05,
max_depth=4,
max_features='sqrt',
min_samples_leaf=15,
min_samples_split=10,
loss='huber',
random_state=5)
GBoost.data = 0
class AveragingModels(BaseEstimator, RegressorMixin, TransformerMixin):
def __init__(self, models):
self.models = models
self.data_mod = [m.data for m in self.models]
self.X = [train.values, train_lasso.values]
self.y = y_train
# we define clones of the original models to fit the data in
def fit(self, X, y):
####dare in imput X = [train.values, train_lasso.values]
self.models_ = [clone(x) for x in self.models]
# Train cloned base models
