def fit(self, X_train, y_train, user_F, user_N, user_M, random_state=None):
"""Fits a random forest classifier to X_train and y_train.
Args:
X_train(list of list of obj): The list of training instances (samples).
The shape of X_train is (n_train_samples, n_features)
y_train(list of obj): The target y values (parallel to X_train)
The shape of y_train is n_train_samples
"""
if random_state is not None:
# store seed
self.random_state = random_state
np.random.seed(self.random_state)
self.X_train = X_train
self.y_train = y_train
self.F = user_F
self.N = user_N
self.M = user_M
stratified_test, stratified_remainder = myevaluation.random_stratified_test_remainder_set(
X_train, y_train, random_state)
train = myutils.stitch_x_and_y_trains(X_train, y_train)
attribute_domains = myutils.calculate_attribute_domains(
train) # TODO: think about if this should be X_train or "train"
N_forest = []
for i in range(self.N):
bootstrapped_table = myutils.bootstrap(stratified_remainder,
random_state)
available_attributes = myutils.get_generic_header(
bootstrapped_table
) # TODO: check that this is used for only X_trains
tree = myutils.tdidt(bootstrapped_table, available_attributes,
attribute_domains, self.F)
N_forest.append(tree)
header = myutils.get_generic_header(stratified_remainder)
header.append("y")
y_predicted = []
y_true = []
all_accuracies = []
# testing accuracy of N_forest trees to find the top M accuracies
for tree in N_forest:
y_predicted_row = []
for item in stratified_test:
y_predicted_row.append(
myutils.tdidt_predict(header, tree, item[:-1]))
y_predicted.append(y_predicted_row)
y_true = myutils.get_column(stratified_test, header, "y")
for predicted_sublist in y_predicted:
accuracy, _ = myutils.accuracy_errorrate(predicted_sublist, y_true)
all_accuracies.append(accuracy)
for _ in range(self.M):
max_ind = all_accuracies.index(max(all_accuracies))
self.forest.append(N_forest[max_ind])
all_accuracies[max_ind] = -1
def random_stratified_test_remainder_set(X, y, random_state, set_size=0.33):
# does not seem random?
if random_state is not None:
# store seed
random_state = random_state
np.random.seed(random_state)
randomize_in_place(X, y)
stitched_table = myutils.stitch_x_and_y_trains(X, y)
header = myutils.get_generic_header(stitched_table)
header.append("y")
names, group_subtables = myutils.group_by_value(stitched_table, header,
"y")
bins = []
for _ in range(3):
bins.append([])
# split data into bins
index = 0
for subtable in group_subtables:
for row in subtable:
bins[index].append(row)
index = (index + 1) % 3
X_train = bins[0]
X_remainder = bins[1] + bins[2]
return X_train, X_remainder
def fit(self, X_train, y_train):
"""Fits a decision tree classifier to X_train and y_train using the TDIDT (top down induction of decision tree) algorithm.
Args:
X_train(list of list of obj): The list of training instances (samples).
The shape of X_train is (n_train_samples, n_features)
y_train(list of obj): The target y values (parallel to X_train)
The shape of y_train is n_train_samples
Notes:
Since TDIDT is an eager learning algorithm, this method builds a decision tree model
from the training data.
Build a decision tree using the nested list representation described in class.
Store the tree in the tree attribute.
Use attribute indexes to construct default attribute names (e.g. "att0", "att1", ...).
"""
##fit() accepts X_train and y_train
# # TODO: calculate the attribute domains dictionary
# # TODO: calculate a header (e.g. ["att0", "att1", ...])
# # my advice: stitch together X_train and y_train
# train = [X_train[i] + [y_train[i]] for i in range(len(X_train))]
# available_attributes = header.copy() # recall: Python is pass
# # by object reference
# # initial tdidt() call
self.X_train = X_train
self.y_train = y_train
train = myutils.stitch_x_and_y_trains(X_train, y_train)
available_attributes = myutils.get_generic_header(
X_train) # TODO: check that this is used for only X_trains
attribute_domains = myutils.calculate_attribute_domains(
X_train) # TODO: think about if this should be X_train or "train"
self.tree = myutils.tdidt(train, available_attributes,
attribute_domains, None)
def predict(self, X_test):
"""Makes predictions for test instances in X_test.
Args:
X_test(list of list of obj): The list of testing samples
The shape of X_test is (n_test_samples, n_features)
Returns:
y_predicted(list of obj): The predicted target y values (parallel to X_test)
"""
train = myutils.stitch_x_and_y_trains(self.X_train, self.y_train)
header = myutils.get_generic_header(train)
y_predicted = []
for item in X_test:
y_predicted.append(myutils.tdidt_predict(header, self.tree, item))
return y_predicted