class _TreeBag:
def __init__(self, features: np.ndarray, data: np.ndarray,
labels: np.ndarray, max_depth: int, min_leaf_points: int):
self.features = features
self.d_tree = Tree(max_depth=max_depth,
min_node_points=min_leaf_points)
self.d_tree.train(data, labels)
def predict(self, x):
x_bag = x[self.features]
return self.d_tree.predict(x_bag)
# declare shared hyper parameters
max_depth = 100
min_samples = 10
tree_clf = DecisionTreeClassifier(max_depth=max_depth, min_samples_leaf=min_samples)
tree_clf.fit(X,y)
sk_tree_graph = export_graphviz(tree_clf,
out_file=None,
class_names=iris.target_names,
rounded=True,
filled=True)
# build tree graph for sklearn model
folder = os.path.dirname(__file__)
sk_tree_source = Source(sk_tree_graph,
filename='skl_treebag.gv',
directory=folder+'\images',
format='png')
sk_tree_source.render()
# my tree class
t = Tree(max_depth=max_depth, min_node_points=min_samples)
Xme = X
Yme = np.atleast_2d(y).T
t.train(Xme, Yme)
build_graph(t, '.\images')
print(t.predict(np.array([1,7])))
import numpy as np from tree.node import Node from tree.tree import Tree x = np.random.random((100, 10)) y = np.atleast_2d(np.random.randint(1, size=100)).T N = Tree() N.train(x, y) x_test = np.random.random((1, 10)) result = N.predict(x=x_test) print(result)
