def test():
with UnitTimer(12):
for _ in range(12):
for i, j in group((x0, x1, x2)):
# print(i.shape, j.shape)
pass
from odin.search import (diagonal_beam_search, diagonal_bruteforce_search,
diagonal_greedy_search, diagonal_hillclimb_search)
from odin.utils import UnitTimer
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
os.environ['TF_FORCE_GPU_ALLOW_GROWTH'] = 'true'
tf.random.set_seed(8)
np.random.seed(8)
shape = (8, 8)
mat = np.random.randint(0, 88, size=shape)
print(mat)
with UnitTimer():
ids = diagonal_beam_search(mat)
print(ids)
print(mat[:, ids])
print(np.sum(np.diag(mat[:, ids])))
with UnitTimer():
ids = diagonal_hillclimb_search(mat)
print(ids)
print(mat[:, ids])
print(np.sum(np.diag(mat[:, ids])))
with UnitTimer():
ids = diagonal_greedy_search(mat)
print(ids)
print(mat[:, ids])
X_valid, y_mRNA_valid, y_prot_valid, y_bin_valid, y_prob_valid = data[
valid], y_mRNA[valid], y_prot[valid], y_bin[valid], y_prob[valid]
X_test, y_mRNA_test, y_prot_test, y_bin_test, y_prob_test = data[test], y_mRNA[
test], y_prot[test], y_bin[test], y_prob[test]
print(ctext("Train:", 'cyan'), X_train.shape, y_mRNA_train.shape,
y_prot_train.shape, y_bin_train.shape, y_prob_train.shape)
print(ctext("Valid:", 'cyan'), X_valid.shape, y_mRNA_valid.shape,
y_prot_valid.shape, y_bin_valid.shape, y_prob_valid.shape)
print(ctext("Test:", 'cyan'), X_test.shape, y_mRNA_test.shape,
y_prot_test.shape, y_bin_test.shape, y_prob_test.shape)
# ===========================================================================
# Evaluation
# ===========================================================================
random_state = get_rng().randint(0, 10e8)
# ====== create transformer ====== #
with UnitTimer(name="Fit PCA"):
pca = PCA(n_components=NUM_COMPONENTS, random_state=random_state)
pca.fit(X_train)
with UnitTimer(name="Fit PPCA"):
ppca = PPCA(n_components=NUM_COMPONENTS,
verbose=True,
random_state=random_state)
ppca.fit(X_train)
with UnitTimer(name="Fit S-PPCA"):
sppca = SupervisedPPCA(n_components=NUM_COMPONENTS,
verbose=True,
extractor='supervised',
random_state=random_state)
X_, y_ = [], []
T.sum(const2.T**3))
outputs, update = theano.scan(step_non,
sequences=theano.shared(
np.arange(12 * 12 * 12 * 8 * 8).reshape(
12 * 12 * 12, 8, 8)),
outputs_info=theano.shared(np.ones((8, 8))),
strict=False)
f_non = theano.function(inputs=[], outputs=outputs, allow_input_downcast=True)
time.sleep(0.5)
for i in range(3):
print('Non-strict scan:')
with UnitTimer(8):
for i in range(8):
f_non()
print('Strict scan:')
with UnitTimer(8):
for i in range(8):
f_strict()
# Non - strict scan:
# Time: 0.064988 (sec)
# Strict scan:
# Time: 0.058314 (sec)
# Non - strict scan:
# Time: 0.059891 (sec)
# Strict scan: