def test_splits():
oob = BootstrapOutOfBag(n_splits=3, random_seed=123)
results = list(oob.split(np.array([1, 2, 3, 4, 5])))
assert len(results) == 3
assert np.array_equal(results[0][0], np.array([2, 4, 2, 1, 3]))
assert np.array_equal(results[0][1], np.array([0]))
assert np.array_equal(results[-1][0], np.array([1, 1, 0, 0, 1]))
assert np.array_equal(results[-1][1], np.array([2, 3, 4]))
def test_splits():
oob = BootstrapOutOfBag(n_splits=3, random_seed=123)
results = list(oob.split(np.array([1, 2, 3, 4, 5])))
assert len(results) == 3
assert np.array_equal(results[0][0], np.array([2, 4, 2, 1, 3]))
assert np.array_equal(results[0][1], np.array([0]))
assert np.array_equal(results[-1][0], np.array([1, 1, 0, 0, 1]))
assert np.array_equal(results[-1][1], np.array([2, 3, 4]))
def get_bootstrap(X: np.ndarray, y: np.ndarray, n: int = 200):
"""Returns train, test, validation splits
Arguments:
X {np.ndarray} -- Feature matrix
y {np.ndarray} -- Label vector
n {int} -- number of bootstrap resamplings
"""
assert len(X) == len(y)
bs = BootstrapOutOfBag(n_splits=n, random_seed=RANDOM_SEED)
oob = bs.split(np.arange(len(y)))
return oob
def test_get_n_splits():
oob = BootstrapOutOfBag(n_splits=3, random_seed=123)
assert oob.n_splits == 3
def test_defaults():
oob = BootstrapOutOfBag()
results = list(oob.split(np.array([1, 2, 3, 4, 5])))
assert len(results) == 200
def test_defaults():
oob = BootstrapOutOfBag()
results = list(oob.split(np.array([1, 2, 3, 4, 5])))
assert len(results) == 200