def test_mtry(self, lung_X, lung_y, mtry):
tree = GRFTreeSurvival(mtry=mtry)
tree.fit(lung_X, lung_y)
if mtry is not None:
assert tree.mtry_ == mtry
else:
assert tree.mtry_ == 3
def test_apply(self, lung_X, lung_y):
tree = GRFTreeSurvival()
tree.fit(lung_X, lung_y)
leaves = tree.apply(lung_X)
assert isinstance(leaves, np.ndarray)
assert np.all(leaves > 0)
assert len(leaves) == len(lung_X)
def test_fit(self, lung_X, lung_y):
tree = GRFTreeSurvival()
with pytest.raises(NotFittedError):
check_is_fitted(tree)
tree.fit(lung_X, lung_y)
check_is_fitted(tree)
assert hasattr(tree, "grf_forest_")
assert hasattr(tree, "mtry_")
assert tree.grf_forest_["num_trees"] == 1
assert tree.criterion == "logrank"
def test_with_X_nan(self, lung_X, lung_y):
lung_X_nan = lung_X.copy()
index = np.random.choice(lung_X_nan.size, 100, replace=False)
lung_X_nan = lung_X_nan.to_numpy()
shape = lung_X_nan.shape
lung_X_nan = lung_X_nan.ravel()
lung_X_nan[index] = np.nan
lung_X_nan = lung_X_nan.reshape(shape)
assert np.sum(np.isnan(lung_X_nan)) == 100
tree = GRFTreeSurvival()
tree.fit(lung_X_nan, lung_y)
pred = tree.predict(lung_X_nan)
assert len(pred) == lung_X_nan.shape[0]
def test_serialize(self, lung_X, lung_y):
tree = GRFTreeSurvival()
# not fitted
tf = tempfile.TemporaryFile()
pickle.dump(tree, tf)
tf.seek(0)
tree = pickle.load(tf)
tree.fit(lung_X, lung_y)
# fitted
tf = tempfile.TemporaryFile()
pickle.dump(tree, tf)
tf.seek(0)
new_tree = pickle.load(tf)
pred = new_tree.predict(lung_X)
assert len(pred) == lung_X.shape[0]
def test_tree_interface(self, lung_X, lung_y):
tree = GRFTreeSurvival()
tree.fit(lung_X, lung_y)
# access attributes the way we would expect to in sklearn
tree_ = tree.tree_
children_left = tree_.children_left
children_right = tree_.children_right
children_default = tree_.children_default
feature = tree_.feature
threshold = tree_.threshold
max_depth = tree_.max_depth
n_node_samples = tree_.n_node_samples
weighted_n_node_samples = tree_.weighted_n_node_samples
node_count = tree_.node_count
capacity = tree_.capacity
n_outputs = tree_.n_outputs
n_classes = tree_.n_classes
value = tree_.value
def get_estimator(self, idx):
"""Extract a single estimator tree from the forest.
:param int idx: The index of the tree to extract.
"""
check_is_fitted(self)
if not self.enable_tree_details:
raise ValueError(
"enable_tree_details must be True prior to training")
return GRFTreeSurvival.from_forest(self, idx=idx)
def test_alpha(self, lung_X, lung_y, alpha):
tree = GRFTreeSurvival(alpha=alpha)
if alpha <= 0 or alpha >= 0.25:
with pytest.raises(ValueError):
tree.fit(lung_X, lung_y)
else:
tree.fit(lung_X, lung_y)
def test_sample_fraction(self, lung_X, lung_y, sample_fraction):
tree = GRFTreeSurvival(sample_fraction=sample_fraction)
if sample_fraction <= 0 or sample_fraction > 1:
with pytest.raises(ValueError):
tree.fit(lung_X, lung_y)
else:
tree.fit(lung_X, lung_y)
def test_decision_path(self, lung_X, lung_y):
tree = GRFTreeSurvival()
tree.quantiles = [0.2]
tree.fit(lung_X, lung_y)
paths = tree.decision_path(lung_X)
assert isinstance(paths, csr_matrix)
assert paths.shape[0] == len(lung_X)
def estimators_(self):
try:
check_is_fitted(self)
except NotFittedError:
raise AttributeError(
f"{self.__class__.__name__} object has no attribute 'estimators_'"
) from None
if not self.enable_tree_details:
raise ValueError(
"enable_tree_details must be True prior to training")
return [
GRFTreeSurvival.from_forest(self, idx=idx)
for idx in range(self.n_estimators)
]
def test_honesty_fraction(self, lung_X, lung_y, honesty_fraction):
tree = GRFTreeSurvival(honesty=True,
honesty_fraction=honesty_fraction,
honesty_prune_leaves=True)
if honesty_fraction <= 0 or honesty_fraction >= 1:
with pytest.raises(RuntimeError):
tree.fit(lung_X, lung_y)
else:
tree.fit(lung_X, lung_y)
def test_equalize_cluster_weights(self, lung_X, lung_y, lung_cluster,
equalize_cluster_weights):
tree = GRFTreeSurvival(
equalize_cluster_weights=equalize_cluster_weights)
tree.fit(lung_X, lung_y, cluster=lung_cluster)
if equalize_cluster_weights:
assert tree.samples_per_cluster_ == 20
else:
assert tree.samples_per_cluster_ == lung_y.shape[0] - 20
if equalize_cluster_weights:
with pytest.raises(ValueError):
tree.fit(
lung_X,
lung_y,
cluster=lung_cluster,
sample_weight=np.ones(lung_y.shape),
)
tree.fit(lung_X, lung_y, cluster=None)
assert tree.samples_per_cluster_ == 0
def test_get_n_leaves(self, lung_X, lung_y):
tree = GRFTreeSurvival()
tree.fit(lung_X, lung_y)
leaves = tree.get_n_leaves()
assert isinstance(leaves, int)
assert np.all(leaves > 0)
def test_clone(self, lung_X, lung_y):
tree = GRFTreeSurvival()
tree.fit(lung_X, lung_y)
clone(tree)
def test_honesty(self, lung_X, lung_y, honesty):
tree = GRFTreeSurvival(honesty=honesty)
tree.fit(lung_X, lung_y)
def test_predict(self, lung_X, lung_y):
tree = GRFTreeSurvival()
tree.fit(lung_X, lung_y)
pred = tree.predict(lung_X)
assert len(pred) == lung_X.shape[0]
def test_honesty_prune_leaves(self, lung_X, lung_y, honesty_prune_leaves):
tree = GRFTreeSurvival(honesty=True,
honesty_prune_leaves=honesty_prune_leaves)
tree.fit(lung_X, lung_y)
def test_get_tags(self):
rfs = GRFTreeSurvival()
tags = rfs._get_tags()
assert tags["requires_y"]
def test_from_forest(self, lung_X, lung_y):
forest = GRFForestSurvival()
forest.fit(lung_X, lung_y)
tree = GRFTreeSurvival.from_forest(forest=forest, idx=0)
tree.predict(lung_X)
def test_get_depth(self, lung_X, lung_y):
tree = GRFTreeSurvival()
tree.fit(lung_X, lung_y)
depth = tree.get_depth()
assert isinstance(depth, int)
assert depth > 0
def test_init(self):
_ = GRFTreeSurvival()
