import numpy as np
from ..tree.tree import ClassifyTree, RegressionTree
from abc import ABCMeta
from scipy.stats import mode
class RandomForest(metaclass=ABCMeta):
"""
Attributes
----------
num_trees : the number of trees to be made in the forest
max_depth : the maximum depth that each tree is allowed to grow
min_size : the minimum number of data observations needed in each split
sample_percentage : size of data to be sampled per tree
Note
----
This class is not to be instantiated. It is simply a base class for the
classification and regression forest classes
"""
def __init__(self,
num_trees,
seed, max_depth,
min_size,
sample_percentage):
"""
Initializes the random forest
Parameters
----------
num_trees : the number of trees to be made in the forest
seed : the seed from which the random sample choices will be made
max_depth : the maximum depth that each tree is allowed to grow
min_size : the minimum number of data observations needed in each split
sample_percentage : size of data to be sampled per tree
"""
self.num_trees = num_trees
self.max_depth = max_depth
self.min_size = min_size
self.sample_percentage = sample_percentage
np.random.seed(seed)
def fit(self, X, y):
"""
Grows a forest of decision trees based off the num_trees
attribute
Parameters
----------
X : N x D matrix of real or ordinal values
y : size N vector consisting of either real values or labels for corresponding
index in X
"""
data = np.column_stack((X, y))
self.forest = np.empty(shape=self.num_trees, dtype='object')
sample_size = int(X.shape[0] * self.sample_percentage)
for i in range(self.num_trees):
sample = data[np.random.choice(data.shape[0], sample_size, replace=True)]
sampled_X = data[:, :data.shape[1] - 1]
sampled_y = data[:, data.shape[1] - 1]
if isinstance(self, RegressionForest):
tree = RegressionTree(
max_depth=self.max_depth,
min_size=self.min_size,
in_forest=True)
else:
tree = ClassifyTree(
max_depth=self.max_depth,
min_size=self.min_size,
in_forest=True)
tree.fit(sampled_X, sampled_y)
self.forest[i] = tree
def predict(self, X):
"""
Predicts the output (y) of a given matrix X
Parameters
----------
X : numerical or ordinal matrix of values corresponding to some output
Returns
-------
The predict values corresponding to the inputs
"""
votes = np.zeros(shape=(self.num_trees, X.shape[0]))
for i, tree in enumerate(self.forest):
votes[i] = tree.predict(X)
predictions = np.zeros(shape=X.shape[0])
if isinstance(self, RegressionForest):
predictions = votes.mean(axis=0)
else:
# print(votes)
predictions = np.squeeze(mode(votes, axis=0)[0])
return predictions
class RegressionForest(RandomForest):
"""
Attributes
----------
num_trees : the number of trees to be made in the forest
max_depth : the maximum depth that each tree is allowed to grow
cost_func : function that determines the cost of each split in the trees
min_size : the minimum number of data observations needed in each split
sample_percentage : size of data to be sampled per tree
"""
def __init__(self,
num_trees=10,
seed=0,
max_depth=None,
min_size=1,
sample_percentage=1):
"""
Initializes Regression Forest
Parameters
----------
num_trees : the number of trees to be made in the forest
seed : the seed from which the random sample choices will be made
max_depth : the maximum depth that each tree is allowed to grow
min_size : the minimum number of data observations needed in each split
sample_percentage : size of data to be sampled per tree
"""
self.num_trees = num_trees
self.max_depth = max_depth
self.min_size = min_size
self.sample_percentage = sample_percentage
super(RegressionForest, self).__init__(
num_trees=num_trees,
seed=seed,
max_depth=max_depth,
min_size=min_size,
sample_percentage=sample_percentage
)
class ClassificationForest(RandomForest):
"""
Attributes
----------
num_trees : the number of trees to be made in the forest
max_depth : the maximum depth that each tree is allowed to grow
cost_func : function that determines the cost of each split in the trees
min_size : the minimum number of data observations needed in each split
sample_percentage : size of data to be sampled per tree
"""
def __init__(self,
num_trees=10,
seed=0,
max_depth=None,
min_size=1,
sample_percentage=1):
"""
Initializes Regression Forest
Parameters
----------
num_trees : the number of trees to be made in the forest
seed : the seed from which the random sample choices will be made
max_depth : the maximum depth that each tree is allowed to grow
cost_func : function that determines the cost of each split in the trees
min_size : the minimum number of data observations needed in each split
sample_percentage : size of data to be sampled per tree
"""
self.num_trees = num_trees
self.max_depth = max_depth
self.min_size = min_size
self.sample_percentage = sample_percentage
super(ClassificationForest,self).__init__(
num_trees=num_trees,
seed=seed,
max_depth=max_depth,
min_size=min_size,
sample_percentage=sample_percentage
)
