def fit(self, X, y):
self.trees = []
for m in range(self.num_trees):
print("Fitting tree %02d/%d..." % (m+1,self.num_trees))
tree = RandomTree(max_depth = self.max_depth)
tree.fit(X,y)
self.trees.append(tree)
def fit(self, X, y):
# Fit each tree
treeList = []
for i in range(self.num_trees):
tree = RandomTree(self.max_depth)
tree.fit(X, y)
treeList.append(tree)
self.treeList = treeList
def fit(self, X, y):
self.random_trees = []
for i in range(self.num_trees):
random_tree = RandomTree(max_depth=self.max_depth)
random_tree.fit(X, y)
self.random_trees.append(random_tree)
def fit(self, X, y):
N, D = X.shape
trees = [None] * self.num_trees
for i in range(self.num_trees):
model = RandomTree(max_depth=self.max_depth)
model.fit(X, y)
trees[i] = model
self.trees = trees
def main():
""" Function creates new tree with all files and directories.
"""
tree = RandomTree(3, 3)
for item in recognize_type(tree.get()):
create(*item)
print(repr(tree))
def fit(self, X, y):
# Train data on each tree
# initialize a forest
self.random_forest = []
for tree in range(self.num_trees):
# Bootstrapping and Random Trees Step
one_tree = RandomTree(max_depth=self.max_depth)
one_tree.fit(X, y)
self.random_forest.append(one_tree)
def fit(self, X, y):
numTrees = self.num_trees
list_of_trees = []
for x in range(numTrees):
tree = RandomTree(max_depth=self.max_depth)
tree.fit(X, y)
list_of_trees.append(tree)
self.list_of_trees = list_of_trees
def fit(self, X, y):
listOfModels = []
for i in range(0, self.num_trees):
model = RandomTree(max_depth=np.inf)
model.fit(X, y)
listOfModels.append(model)
self.LOM = listOfModels
self.y_length = y.shape[0]
def fit(self, X, y):
self.trees = []
num_trees = self.num_trees
max_depth = self.max_depth
for m in range(num_trees):
tree = RandomTree(max_depth=max_depth)
tree.fit(X, y)
self.trees.append(tree)
def load(self, filename):
file = open(filename, "r")
jsonModel = file.read()
model = json.loads(jsonModel)["model"]
self.trees = []
for i in range(len(model)):
tree = RandomTree(max_depth=self.max_depth)
tree.load(model[i])
self.trees.append(tree)
file.close()
def fit(self, X, y):
forest = []
for n in range(self.num_trees):
model = RandomTree( max_depth = self.max_depth)
model.fit(X,y)
forest.append(model)
self.forest = forest
def fit(self, X, y):
self.X = X
self.y = y
self.stats = []
for i in range(self.num_trees):
self.stats = np.append(self.stats, RandomTree(self.max_depth))
self.stats[i].fit(X, y)
def fit(self, X, y):
rt = []
M = self.num_trees
for n in range(M):
#rt.append(RandomTree(self.max_depth))
rt.append(RandomTree.fit(self, X, y))
self.rt = rt
def __init__(self, num_trees, max_depth):
self.num_trees = num_trees
self.max_depth = max_depth
# Create array of size num_trees with RandomTree models
self.rand_trees = []
for i in range(self.num_trees):
self.rand_trees.append(RandomTree(max_depth=self.max_depth))
def fit(self, X, Y):
n, d = X.shape
self.trees = []
for i in range(self.n_trees):
idx = np.arange(n)
np.random.seed(np.int(time() / 150))
np.random.shuffle(idx)
X = X[idx]
Y = Y[idx]
train = np.int(self.ratio_per_tree * n)
Xtrain = X[:train, :]
Ytrain = Y[:train]
clf = RandomTree(max_depth=self.max_depth,
ratio_features=self.ratio_features)
clf.fit(Xtrain, Ytrain)
self.trees.append(clf)
def gen_random_forest(dataset,
attributes,
ntrees,
nattributes,
depth_limit=None):
"""
:param dataset: the dataset dataframe
:param attributes: the attributes dict
:param ntrees: int, number of trees
:param nattributes: int, number of attributes of each tree (aka. "m")
Algorithm in slide 32 class 16
"""
return [
RandomTree(*select_attributes(bootstrap(dataset), attributes,
nattributes),
depth_limit=depth_limit) for _ in range(0, ntrees)
]
if question == '5':
dataset = utils.load_dataset('vowel')
X = dataset['X']
y = dataset['y']
X_test = dataset['Xtest']
y_test = dataset['ytest']
print("n = %d, d = %d" % X.shape)
def evaluate_model(model):
model.fit(X, y)
y_pred = model.predict(X)
tr_error = np.mean(y_pred != y)
t = time.time()
y_pred = model.predict(X_test)
te_error = np.mean(y_pred != y_test)
print("Training error: %.3f" % tr_error)
print("Testing error: %.3f" % te_error)
print("Time taken: %f" % (time.time() - t))
evaluate_model(
DecisionTree(max_depth=50, stump_class=DecisionStumpInfoGain))
evaluate_model(RandomTree(max_depth=50))
evaluate_model(
RandomForestClassifier(n_estimators=50,
max_depth=50,
criterion='entropy',
random_state=1))
evaluate_model(RandomForest(num_trees=50, max_depth=np.inf))
def fit(self, X, y):
self.Random_Forest = []
for x in range(0, self.num_trees):
New_tree = RandomTree(self.max_depth)
New_tree.fit(X, y)
self.Random_Forest.append(New_tree)
def __init__(self, max_depth, num_trees):
RandomTree.__init__(self, max_depth)
self.num_trees = num_trees
model.fit(X,y)
y_pred = model.predict(X)
tr_error = np.mean(y_pred != y)
y_pred = model.predict(X_test)
te_error = np.mean(y_pred != y_test)
print(" Training error: %.3f" % tr_error)
print(" Testing error: %.3f" % te_error)
print("Our implementations:")
print(" Decision tree info gain")
evaluate_model(DecisionTree(max_depth=np.inf, stump_class=DecisionStumpInfoGain))
print(" Random tree info gain")
evaluate_model(RandomTree(max_depth=np.inf))
print(" Random forest info gain")
evaluate_model(RandomForest(max_depth=np.inf, num_trees=50))
print("sklearn implementations")
print(" Decision tree info gain")
evaluate_model(DecisionTreeClassifier(criterion="entropy"))
print(" Random forest info gain")
evaluate_model(RandomForestClassifier(criterion="entropy"))
print(" Random forest info gain, more trees")
evaluate_model(RandomForestClassifier(criterion="entropy", n_estimators=50))
elif question == '3':
X = load_dataset('clusterData.pkl')['X']
def fit(self, X, y):
for i in range(self.num_trees):
model = RandomTree(max_depth=self.max_depth)
model.fit(X, y)
self.models.append(model)
mode_parser = parser()
if str(mode_parser.mode) == 'verify':
# benchmark
m = []
with open('datasets/benchmark/benchmark.csv') as bfile:
breader = csv.reader(bfile, delimiter=';')
for row in breader:
m.append(row)
attributes = {x: ['categorical'] for x in m[0][:-1]}
attributes_names = m[0][:-1]
del m[0]
dataset = pd.DataFrame(m, columns=attributes_names + ['y'])
for x in attributes.keys():
attributes[x].append(dataset[x].unique())
RT = RandomTree(dataset, attributes)
RT.print_tree()
sys.exit()
elif str(mode_parser.mode) == 'wine':
dataset, attributes = read_dataset('wine')
elif str(mode_parser.mode) == 'survival':
dataset, attributes = read_dataset('survival')
elif str(mode_parser.mode) == 'cancer':
dataset, attributes = read_dataset('cancer')
elif str(mode_parser.mode) == 'contraceptive':
dataset, attributes = read_dataset('contraceptive')
"""
def __init__(self, max_depth, num_trees):
self.num_trees = num_trees
RandomTree.__init__(self, max_depth=max_depth)
DecisionTree.__init__(self,
max_depth=max_depth,
stump_class=RandomStumpInfoGain)