def test_original_biased_nonlin_semi_nmf(self):
auv = sio.loadmat(mat_file)
u, v = auv['u'], auv['v']
a = relu(u @ v)
bias_v = np.vstack((v, np.ones((1, v.shape[1]))))
old_loss = np_frobenius_norm(a, u @ v)
a_ph = tf.placeholder(tf.float64, shape=a.shape)
u_ph = tf.placeholder(tf.float64, shape=u.shape)
bias_v_ph = tf.placeholder(tf.float64, shape=bias_v.shape)
tf_bias_u, tf_v = nonlin_semi_nmf(a_ph, u_ph, bias_v_ph, use_bias=True, use_tf=True, num_calc_v=0)
init = tf.global_variables_initializer()
with tf.Session() as sess:
init.run()
start_time = time.time()
_u, _bias_v = sess.run([tf_bias_u, tf_v], feed_dict={a_ph: a, u_ph: u, bias_v_ph: bias_v})
end_time = time.time()
duration = end_time - start_time
_bias_u = np.hstack((_u, np.ones((_u.shape[0], 1))))
new_loss = np_frobenius_norm(a, relu(_bias_u @ _bias_v))
assert a.shape == (_bias_u @ _bias_v).shape
assert new_loss < old_loss, "new loss should be less than old loss."
print_format('TensorFlow', 'biased Nonlinear semi-NMF(NOT CALC v)', a, _bias_u, _bias_v, old_loss, new_loss,
duration)
def test_np_not_calc_v_biased_nonlin_semi_nmf(self):
auv = sio.loadmat(mat_file)
a, u, v = auv['a'], auv['u'], auv['v']
old_loss = np_frobenius_norm(a, u @ v)
biased_u = np.hstack((u, np.ones((u.shape[0], 1))))
start_time = time.time()
biased_u, v = nonlin_semi_nmf(a,
biased_u,
v,
use_bias=True,
num_calc_v=0)
end_time = time.time()
duration = end_time - start_time
bias_v = np.vstack((v, np.ones((1, v.shape[1]))))
new_loss = np_frobenius_norm(a, relu(biased_u @ bias_v))
assert a.shape == (biased_u @ bias_v).shape
assert new_loss < old_loss, "new loss should be less than old loss."
print('\n[Numpy]Solve biased Nonlinear semi-NMF(NOT CALCULATE v)\n\t'
'old loss {0}\n\t'
'new loss {1}\n\t'
'process duration {2}'.format(old_loss, new_loss, duration))
def test_np_not_calc_v_vanilla_nonlin_semi_nmf(self):
a = np.random.uniform(0., 1., size=(100, 100))
u = np.random.uniform(0., 1., size=(100, 300))
v = np.random.uniform(-1., 1., size=(300, 100))
old_loss = np_frobenius_norm(a, u @ v)
start_time = time.time()
u, v = nonlin_semi_nmf(a, u, v, use_bias=False, num_calc_v=0)
assert np.min(u) > 0, np.min(u)
end_time = time.time()
duration = end_time - start_time
new_loss = np_frobenius_norm(a, relu(u @ v))
assert a.shape == (u @ v).shape
assert new_loss < old_loss, "new loss should be less than old loss."
print_format('Numpy', 'Nonlinear semi-NMF(NOT CALCULATE v)', a, u, v, old_loss, new_loss, duration)
def test_np_not_calc_v_vanilla_nonlin_semi_nmf(self):
auv = sio.loadmat(mat_file)
a, u, v = auv['a'], auv['u'], auv['v']
old_loss = np_frobenius_norm(a, u @ v)
start_time = time.time()
u, v = nonlin_semi_nmf(a, u, v, use_bias=False, num_calc_v=0)
end_time = time.time()
duration = end_time - start_time
new_loss = np_frobenius_norm(a, relu(u @ v))
assert a.shape == (u @ v).shape
assert new_loss < old_loss, "new loss should be less than old loss."
print('\n[Numpy]Solve Nonlinear semi-NMF(NOT CALCULATE v)\n\t'
'old loss {0}\n\t'
'new loss {1}\n\t'
'process duration {2}'.format(old_loss, new_loss, duration))
def test_np_not_calc_v_biased_nonlin_semi_nmf(self):
a = np.random.uniform(0., 1., size=(100, 100))
u = np.random.uniform(0., 1., size=(100, 300))
v = np.random.uniform(-1., 1., size=(300, 100))
old_loss = np_frobenius_norm(a, u @ v)
bias_v = np.vstack((v, np.ones((1, v.shape[1]))))
start_time = time.time()
u, bias_v = nonlin_semi_nmf(a, u, bias_v, use_bias=True, num_calc_v=0)
assert np.min(u) > 0, np.min(u)
end_time = time.time()
duration = end_time - start_time
bias_u = np.hstack((u, np.ones((u.shape[0], 1))))
new_loss = np_frobenius_norm(a, relu(bias_u @ bias_v))
assert a.shape == (bias_u @ bias_v).shape
assert new_loss < old_loss, "new loss should be less than old loss."
print_format('Numpy', 'biased Nonlinear semi-NMF(NOT CALC v)', a, bias_u, bias_v, old_loss, new_loss, duration)
def test_tf_not_calc_v_biased_nonlin_semi_nmf(self):
auv = sio.loadmat(mat_file)
a, u, v = auv['a'], auv['u'], auv['v']
bias_u = np.hstack((u, np.ones((u.shape[0], 1))))
old_loss = np_frobenius_norm(a, u @ v)
a_ph = tf.placeholder(tf.float64, shape=a.shape)
bias_u_ph = tf.placeholder(tf.float64, shape=bias_u.shape)
v_ph = tf.placeholder(tf.float64, shape=v.shape)
tf_bias_u, tf_v = nonlin_semi_nmf(a_ph,
bias_u_ph,
v_ph,
num_calc_v=0,
use_bias=True,
use_tf=True)
init = tf.global_variables_initializer()
with tf.Session() as sess:
init.run()
start_time = time.time()
_bias_u, _v = sess.run([tf_bias_u, tf_v],
feed_dict={
a_ph: a,
bias_u_ph: bias_u,
v_ph: v
})
end_time = time.time()
duration = end_time - start_time
_bias_v = np.vstack((_v, np.ones((1, v.shape[1]))))
new_loss = np_frobenius_norm(a, relu(_bias_u @ _bias_v))
assert a.shape == (_bias_u @ _bias_v).shape
assert new_loss < old_loss, "new loss should be less than old loss."
print(
'\n[TensorFlow]Solve biased Nonlinear semi-NMF(NOT CALCULATE v)\n\t'
'old loss {0}\n\t'
'new loss {1}\n\t'
'process duration {2}'.format(old_loss, new_loss, duration))