def _make_tree(sf):
model = tc.decision_tree_classifier.create(sf,
'target',
validation_set=None,
max_depth=10)
tree = DecisionTree.from_model(model)
return tree
def setUpClass(self):
sf = tc.SFrame({
'cat1': ['1', '1', '2', '2', '2'] * 100,
'cat2': ['1', '3', '3', '1', '1'] * 100,
'target': ['1', '2', '1', '2', '1'] * 100,
})
model = tc.classifier.boosted_trees_classifier.create(
sf, 'target', validation_set=None, max_depth=2)
tree = DecisionTree.from_model(model)
self.tree = tree
def setUpClass(self):
sf = tc.SFrame({
"cat1": ["1", "1", "2", "2", "2"] * 100,
"cat2": ["1", "3", "3", "1", "1"] * 100,
"target": ["1", "2", "1", "2", "1"] * 100,
})
model = tc.classifier.boosted_trees_classifier.create(
sf, "target", validation_set=None, max_depth=2)
tree = DecisionTree.from_model(model)
self.tree = tree
def _run_test(self, sf):
sf['target'] = [i < sf.num_rows()/2 for i in range(sf.num_rows())]
for model in [
tc.regression.boosted_trees_regression,
tc.classifier.boosted_trees_classifier,
tc.regression.random_forest_regression,
tc.classifier.random_forest_classifier,
tc.regression.decision_tree_regression,
tc.classifier.decision_tree_classifier]:
m = model.create(sf, 'target', validation_set = None, max_depth=2)
tree = DecisionTree.from_model(m)
for nid, node in tree.nodes.items():
val = tree.get_prediction_score(nid)
if node.is_leaf:
self.assertTrue(type(val) in {float, int})
else:
self.assertEqual(val, None)