def test_gradient_descent():
t1 = np.array([-5.21e-16, -7.86e-02, 1.02e+00])
ada = Adaline(epochs=30, eta=0.01, learning='gd', random_seed=1)
ada.fit(X_std, y1)
np.testing.assert_almost_equal(ada.w_, t1, 2)
assert((y1 == ada.predict(X_std)).all())
def test_0_1_class():
t1 = np.array([0.51, -0.04, 0.51])
ada = Adaline(epochs=30, eta=0.01, learning='sgd', random_seed=1)
ada.fit(X_std, y0)
np.testing.assert_almost_equal(ada.w_, t1, 2)
assert((y0 == ada.predict(X_std)).all())
def test_stochastic_gradient_descent():
t1 = np.array([0.03, -0.09, 1.02])
ada = Adaline(epochs=30, eta=0.01, learning='sgd', random_seed=1)
ada.fit(X_std, y1)
np.testing.assert_almost_equal(ada.w_, t1, 2)
assert((y1 == ada.predict(X_std)).all())
def test_print_progress_2():
ada = Adaline(epochs=30,
eta=0.01,
minibatches=1,
print_progress=2,
random_seed=1)
ada.fit(X_std, y1)
def test_score_function():
ada = Adaline(epochs=30,
eta=0.01,
minibatches=1,
random_seed=1)
ada.fit(X_std, y1)
acc = ada.score(X_std, y1)
assert acc == 1.0, acc
def test_ary_persistency_in_shuffling():
orig = X_std.copy()
ada = Adaline(epochs=30,
eta=0.01,
minibatches=len(y),
random_seed=1)
ada.fit(X_std, y1)
np.testing.assert_almost_equal(orig, X_std, 6)
def test_invalid_class():
ada = Adaline(epochs=40, eta=0.01, random_seed=1)
try:
ada.fit(X, y2) # 0, 1 class
assert(1==2)
except ValueError:
pass
def test_score_function():
t1 = np.array([-5.21e-16, -7.86e-02, 1.02e+00])
ada = Adaline(epochs=30,
eta=0.01,
minibatches=1,
random_seed=1)
ada.fit(X_std, y1)
acc = ada.score(X_std, y1)
assert acc == 1.0, acc
def test_normal_equation():
t1 = np.array([-5.21e-16, -7.86e-02, 1.02e+00])
ada = Adaline(epochs=30,
eta=0.01,
minibatches=None,
random_seed=1)
ada.fit(X_std, y1)
np.testing.assert_almost_equal(ada.w_, t1, 2)
assert((y1 == ada.predict(X_std)).all())
def test_stochastic_gradient_descent():
t1 = np.array([[-0.08], [1.02]])
ada = Adaline(epochs=30,
eta=0.01,
minibatches=len(y),
random_seed=1)
ada.fit(X_std, y1)
np.testing.assert_almost_equal(ada.w_, t1, 2)
assert((y1 == ada.predict(X_std)).all())
def test_standardized_iris_data_with_shuffle():
t1 = np.array([-5.21e-16, -7.86e-02, 1.02e+00])
ada = Adaline(epochs=30,
eta=0.01,
solver='gd',
random_seed=1,
shuffle=True)
ada.fit(X_std, y1)
np.testing.assert_almost_equal(ada.w_, t1, 2)
assert((y1 == ada.predict(X_std)).all())
def test_standardized_iris_data_with_zero_weights():
t1 = np.array([-5.21e-16, -7.86e-02, 1.02e+00])
ada = Adaline(epochs=30,
eta=0.01,
minibatches=1,
random_seed=1,
zero_init_weight=True)
ada.fit(X_std, y1)
np.testing.assert_almost_equal(ada.w_, t1, 2)
assert((y1 == ada.predict(X_std)).all())
def test_gradient_descent():
t1 = np.array([[-0.08], [1.02]])
b1 = np.array([0.00])
ada = Adaline(epochs=30,
eta=0.01,
minibatches=1,
random_seed=1)
ada.fit(X_std, y1)
np.testing.assert_almost_equal(ada.w_, t1, decimal=2)
np.testing.assert_almost_equal(ada.b_, b1, decimal=2)
assert((y1 == ada.predict(X_std)).all())
def test_normal_equation():
t1 = np.array([[-0.08], [1.02]])
b1 = np.array([0.00])
ada = Adaline(epochs=30,
eta=0.01,
minibatches=None,
random_seed=None)
ada.fit(X_std, y1)
np.testing.assert_almost_equal(ada.w_, t1, decimal=2)
np.testing.assert_almost_equal(ada.b_, b1, decimal=2)
assert (y1 == ada.predict(X_std)).all(), ada.predict(X_std)
def test_refit_weights():
t1 = np.array([-5.21e-16, -7.86e-02, 1.02e+00])
ada = Adaline(epochs=15, eta=0.01, solver='gd', random_seed=1)
ada.fit(X_std, y1, init_weights=True)
ada.fit(X_std, y1, init_weights=False)
np.testing.assert_almost_equal(ada.w_, t1, 2)
assert((y1 == ada.predict(X_std)).all())
def test_refit_weights():
t1 = np.array([[-0.08], [1.02]])
ada = Adaline(epochs=15,
eta=0.01,
minibatches=1,
random_seed=1)
ada.fit(X_std, y1, init_params=True)
ada.fit(X_std, y1, init_params=False)
np.testing.assert_almost_equal(ada.w_, t1, 2)
assert((y1 == ada.predict(X_std)).all())
def test_ary_persistency_in_shuffling():
orig = X_std.copy()
ada = Adaline(epochs=30, eta=0.01, minibatches=len(y), random_seed=1)
ada.fit(X_std, y1)
np.testing.assert_almost_equal(orig, X_std, 6)
from mlxtend.classifier import Perceptron
from mlxtend.classifier import Adaline
from mlxtend.classifier import MultiLayerPerceptron
import numpy as np
import matplotlib.pyplot as plt
from sklearn.datasets import make_moons
import matplotlib.gridspec as gridspec
import itertools
gs = gridspec.GridSpec(2, 2) #xw
X, y = make_moons(n_samples=100, random_state=123)
fig = plt.figure(figsize=(10, 8))
ppn = Perceptron(epochs=50, eta=0.05, random_seed=0)
ppn.fit(X, y)
ada = Adaline(epochs=50, eta=0.05, random_seed=0)
ada.fit(X, y)
mlp = MultiLayerPerceptron(n_output=len(np.unique(y)),
n_features=X.shape[1],
n_hidden=150,
l2=0.0,
l1=0.0,
epochs=500,
eta=0.01,
alpha=0.0,
decrease_const=0.0,
minibatches=1,
shuffle_init=False,
shuffle_epoch=False,
random_seed=0)
def test_invalid_class():
ada = Adaline(epochs=40, eta=0.01, random_seed=1)
try:
ada.fit(X, y2) # 0, 1 class
except ValueError:
pass
def test_score_function():
ada = Adaline(epochs=30, eta=0.01, minibatches=1, random_seed=1)
ada.fit(X_std, y1)
acc = ada.score(X_std, y1)
assert acc == 1.0, acc
def test_gradient_descent():
t1 = np.array([-5.21e-16, -7.86e-02, 1.02e+00])
ada = Adaline(epochs=30, eta=0.01, minibatches=1, random_seed=1)
ada.fit(X_std, y1)
np.testing.assert_almost_equal(ada.w_, t1, 2)
assert ((y1 == ada.predict(X_std)).all())
def test_0_1_class():
t1 = np.array([0.51, -0.04, 0.51])
ada = Adaline(epochs=30, eta=0.01, minibatches=1, random_seed=1)
ada.fit(X_std, y0)
np.testing.assert_almost_equal(ada.w_, t1, 2)
assert ((y0 == ada.predict(X_std)).all())
def test_array_dimensions():
ada = Adaline(epochs=15, eta=0.01, random_seed=1)
ada = ada.fit(np.array([1, 2, 3]), [-1])
def test_array_dimensions():
ada = Adaline(epochs=15, eta=0.01, random_seed=1)
ada = ada.fit(np.array([1, 2, 3]), [-1])
def test_invalid_solver():
t1 = np.array([-5.21e-16, -7.86e-02, 1.02e+00])
ada = Adaline(epochs=30, eta=0.01, solver='bla', random_seed=1)
ada.fit(X_std, y1)
def test_stochastic_gradient_descent():
t1 = np.array([[-0.08], [1.02]])
ada = Adaline(epochs=30, eta=0.01, minibatches=len(y), random_seed=1)
ada.fit(X_std, y1)
np.testing.assert_almost_equal(ada.w_, t1, 2)
assert ((y1 == ada.predict(X_std)).all())
from mlxtend.evaluate import plot_decision_regions
from mlxtend.classifier import Adaline
import matplotlib.pyplot as plt
X, y = iris_data()
X = X[:, [0, 3]]
X = X[0:100]
y = y[0:100]
X[:,0] = (X[:,0] - X[:,0].mean()) / X[:,0].std()
X[:,1] = (X[:,1] - X[:,1].mean()) / X[:,1].std()
# Closed Form Solution
ada = Adaline(epochs=30,
eta=0.01,
minibatches=None,
random_seed=1)
ada.fit(X, y)
plot_decision_regions(X, y, clf=ada)
plt.title('Adaline - Stochastic Gradient Descent')
plt.show()
# (Stochastic) Gradient Descent
ada2 = Adaline(epochs=30,
eta=0.01,
minibatches=1, # 1 for GD learning
#minibatches=len(y), # len(y) for SGD learning
#minibatches=5, # for SGD learning w. minibatch size 20
def test_invalid_labels_2():
y2 = np.where(y1 == 0, -1, 1)
ada = Adaline(epochs=15, eta=0.01, random_seed=1)
assert_raises(AttributeError,
'y array must not contain negative labels.\nFound [-1 1]',
ada.fit, X, y2, {(-1, 1)})