random_forest.py

# coding:utf-8
importnumpyasnp

frommla.baseimportBaseEstimator
frommla.ensemble.baseimportinformation_gain, mse_criterion
frommla.ensemble.treeimportTree


classRandomForest(BaseEstimator):
def__init__(self, n_estimators=10, max_features=None, min_samples_split=10, max_depth=None, criterion=None):
"""Base class for RandomForest.

 Parameters
 ----------
 n_estimators : int
 The number of decision tree.
 max_features : int
 The number of features to consider when looking for the best split.
 min_samples_split : int
 The minimum number of samples required to split an internal node.
 max_depth : int
 Maximum depth of the tree.
 criterion : str
 The function to measure the quality of a split.
 """
self.max_depth=max_depth
self.min_samples_split=min_samples_split
self.max_features=max_features
self.n_estimators=n_estimators
self.trees= []

deffit(self, X, y):
self._setup_input(X, y)
ifself.max_featuresisNone:
self.max_features=int(np.sqrt(X.shape[1]))
else:
assertX.shape[1] >self.max_features
self._train()

def_train(self):
fortreeinself.trees:
tree.train(
self.X,
self.y,
max_features=self.max_features,
min_samples_split=self.min_samples_split,
max_depth=self.max_depth
 )

def_predict(self, X=None):
raiseNotImplementedError()


classRandomForestClassifier(RandomForest):
def__init__(self, n_estimators=10, max_features=None, min_samples_split=10, max_depth=None, criterion="entropy"):
super(RandomForestClassifier, self).__init__(
n_estimators=n_estimators,
max_features=max_features,
min_samples_split=min_samples_split,
max_depth=max_depth,
criterion=criterion,
 )

ifcriterion=="entropy":
self.criterion=information_gain
else:
raiseValueError()

# Initialize empty trees
for_inrange(self.n_estimators):
self.trees.append(Tree(criterion=self.criterion))

def_predict(self, X=None):
y_shape=np.unique(self.y).shape[0]
predictions=np.zeros((X.shape[0], y_shape))

foriinrange(X.shape[0]):
row_pred=np.zeros(y_shape)
fortreeinself.trees:
row_pred+=tree.predict_row(X[i, :])

row_pred/=self.n_estimators
predictions[i, :] =row_pred
returnpredictions


classRandomForestRegressor(RandomForest):
def__init__(self, n_estimators=10, max_features=None, min_samples_split=10, max_depth=None, criterion="mse"):
super(RandomForestRegressor, self).__init__(
n_estimators=n_estimators,
max_features=max_features,
min_samples_split=min_samples_split,
max_depth=max_depth,
 )

ifcriterion=="mse":
self.criterion=mse_criterion
else:
raiseValueError()

# Initialize empty regression trees
for_inrange(self.n_estimators):
self.trees.append(Tree(regression=True, criterion=self.criterion))

def_predict(self, X=None):
predictions=np.zeros((X.shape[0], self.n_estimators))
fori, treeinenumerate(self.trees):
predictions[:, i] =tree.predict(X)
returnpredictions.mean(axis=1)