def test_fit_genetic_alg():
"""Test fit method using the genetic_alg algorithm"""
network = LinearRegression(algorithm='genetic_alg',
bias=False,
learning_rate=1,
clip_max=1,
max_attempts=100)
X = np.array([[0, 1, 0, 1], [0, 0, 0, 0], [1, 1, 1, 1], [1, 1, 1, 1],
[0, 0, 1, 1], [1, 0, 0, 0]])
y = np.reshape(np.array([1, 1, 0, 0, 1, 1]), [6, 1])
network.fit(X, y)
fitted = network.fitted_weights
assert (sum(fitted) < 4 and len(fitted) == 4 and min(fitted) >= -1
and max(fitted) <= 1)
def test_fit_gradient_descent():
"""Test fit method using the gradient_descent algorithm"""
network = LinearRegression(algorithm='gradient_descent',
bias=False,
learning_rate=0.1,
clip_max=1,
max_attempts=100)
X = np.array([[0, 1, 0, 1], [0, 0, 0, 0], [1, 1, 1, 1], [1, 1, 1, 1],
[0, 0, 1, 1], [1, 0, 0, 0]])
y = np.reshape(np.array([1, 1, 0, 0, 1, 1]), [6, 1])
weights = np.array([1, 1, 1, 1])
network.fit(X, y, init_weights=weights)
fitted = network.fitted_weights
assert (sum(fitted) <= 4 and len(fitted) == 4 and min(fitted) >= -1
and max(fitted) <= 1)
def test_predict_bias():
"""Test predict method with bias term"""
network = LinearRegression(algorithm='random_hill_climb',
bias=True,
learning_rate=1,
clip_max=1,
max_attempts=100)
X = np.array([[0, 1, 0, 1], [0, 0, 0, 0], [1, 1, 1, 1], [1, 1, 1, 1],
[0, 0, 1, 1], [1, 0, 0, 0]])
network.fitted_weights = np.array([1, 1, 1, 1, 1])
network.node_list = [5, 1]
network.output_activation = identity
x = np.reshape(np.array([3, 1, 5, 5, 3, 2]), [6, 1])
assert np.array_equal(network.predict(X), x)