def test_check_consistent_shape(self):
X_train, y_train, X_test, y_test = generate_data(
n_train=self.n_train,
n_test=self.n_test,
contamination=self.contamination)
X_train_n, y_train_n, X_test_n, y_test_n, y_train_pred_n, y_test_pred_n \
= check_consistent_shape(X_train, y_train, X_test, y_test,
y_train, y_test)
assert_allclose(X_train_n, X_train)
assert_allclose(y_train_n, y_train)
assert_allclose(X_test_n, X_test)
assert_allclose(y_test_n, y_test)
assert_allclose(y_train_pred_n, y_train)
assert_allclose(y_test_pred_n, y_test)
# test shape difference
with assert_raises(ValueError):
check_consistent_shape(X_train, y_train, y_train, y_test, y_train,
y_test)
# test shape difference between X_train and X_test
X_test = np.hstack((X_test, np.zeros(
(X_test.shape[0], 1)))) # add extra column/feature
with assert_raises(ValueError):
check_consistent_shape(X_train, y_train, X_test, y_test,
y_train_pred_n, y_test_pred_n)
def test_data_generate2(self):
X_train, y_train, X_test, y_test = \
generate_data(n_train=self.n_train,
n_test=self.n_test,
n_features=3,
contamination=self.contamination)
assert_allclose(X_train.shape, (self.n_train, 3))
assert_allclose(X_test.shape, (self.n_test, 3))
def test_data_generate2(self):
X_train, y_train, X_test, y_test = \
generate_data(n_train=self.n_train,
n_test=self.n_test,
n_features=3,
contamination=self.contamination)
assert_allclose(X_train.shape, (self.n_train, 3))
assert_allclose(X_test.shape, (self.n_test, 3))
def test_data_generate3(self):
X_train, y_train, X_test, y_test = \
generate_data(n_train=self.n_train,
n_test=self.n_test,
n_features=2,
contamination=self.contamination,
random_state=42)
X_train2, y_train2, X_test2, y_test2 = \
generate_data(n_train=self.n_train,
n_test=self.n_test,
n_features=2,
contamination=self.contamination,
random_state=42)
assert_allclose(X_train, X_train2)
assert_allclose(X_test, X_test2)
assert_allclose(y_train, y_train2)
assert_allclose(y_test, y_test2)
def test_data_generate3(self):
X_train, y_train, X_test, y_test = \
generate_data(n_train=self.n_train,
n_test=self.n_test,
n_features=2,
contamination=self.contamination,
random_state=42)
X_train2, y_train2, X_test2, y_test2 = \
generate_data(n_train=self.n_train,
n_test=self.n_test,
n_features=2,
contamination=self.contamination,
random_state=42)
assert_allclose(X_train, X_train2)
assert_allclose(X_test, X_test2)
assert_allclose(y_train, y_train2)
assert_allclose(y_test, y_test2)
def test_get_outliers_inliers(self):
X_train, y_train = generate_data(n_train=self.n_train,
train_only=True,
contamination=self.contamination)
X_outliers, X_inliers = get_outliers_inliers(X_train, y_train)
inlier_index = int(self.n_train * (1 - self.contamination))
assert_allclose(X_train[0:inlier_index, :], X_inliers)
assert_allclose(X_train[inlier_index:, :], X_outliers)
def test_get_outliers_inliers(self):
X_train, y_train = generate_data(
n_train=self.n_train, train_only=True,
contamination=self.contamination)
X_outliers, X_inliers = get_outliers_inliers(X_train, y_train)
inlier_index = int(self.n_train * (1 - self.contamination))
assert_allclose(X_train[0:inlier_index, :], X_inliers)
assert_allclose(X_train[inlier_index:, :], X_outliers)
def test_data_generate(self):
X_train, y_train, X_test, y_test = \
generate_data(n_train=self.n_train,
n_test=self.n_test,
contamination=self.contamination)
assert_equal(y_train.shape[0], X_train.shape[0])
assert_equal(y_test.shape[0], X_test.shape[0])
assert_less_equal(self.n_train - X_train.shape[0], 1)
assert_equal(X_train.shape[1], 2)
assert_less_equal(self.n_test - X_test.shape[0], 1)
assert_equal(X_test.shape[1], 2)
out_perc = np.sum(y_train) / self.n_train
assert_allclose(self.contamination, out_perc, atol=0.01)
out_perc = np.sum(y_test) / self.n_test
assert_allclose(self.contamination, out_perc, atol=0.01)
def test_data_generate(self):
X_train, y_train, X_test, y_test = \
generate_data(n_train=self.n_train,
n_test=self.n_test,
contamination=self.contamination)
assert_equal(y_train.shape[0], X_train.shape[0])
assert_equal(y_test.shape[0], X_test.shape[0])
assert_less_equal(self.n_train - X_train.shape[0], 1)
assert_equal(X_train.shape[1], 2)
assert_less_equal(self.n_test - X_test.shape[0], 1)
assert_equal(X_test.shape[1], 2)
out_perc = np.sum(y_train) / self.n_train
assert_allclose(self.contamination, out_perc, atol=0.01)
out_perc = np.sum(y_test) / self.n_test
assert_allclose(self.contamination, out_perc, atol=0.01)
def test_check_consistent_shape(self):
X_train, y_train, X_test, y_test = generate_data(
n_train=self.n_train,
n_test=self.n_test,
contamination=self.contamination)
X_train_n, y_train_n, X_test_n, y_test_n, y_train_pred_n, y_test_pred_n \
= check_consistent_shape(X_train, y_train, X_test, y_test,
y_train, y_test)
assert_allclose(X_train_n, X_train)
assert_allclose(y_train_n, y_train)
assert_allclose(X_test_n, X_test)
assert_allclose(y_test_n, y_test)
assert_allclose(y_train_pred_n, y_train)
assert_allclose(y_test_pred_n, y_test)
# test shape difference
with assert_raises(ValueError):
check_consistent_shape(X_train, y_train, y_train, y_test, y_train,
y_test)
def test_check_consistent_shape(self):
X_train, y_train, X_test, y_test = generate_data(
n_train=self.n_train,
n_test=self.n_test,
contamination=self.contamination)
X_train_n, y_train_n, X_test_n, y_test_n, y_train_pred_n, y_test_pred_n \
= check_consistent_shape(X_train, y_train, X_test, y_test,
y_train, y_test)
assert_allclose(X_train_n, X_train)
assert_allclose(y_train_n, y_train)
assert_allclose(X_test_n, X_test)
assert_allclose(y_test_n, y_test)
assert_allclose(y_train_pred_n, y_train)
assert_allclose(y_test_pred_n, y_test)
# test shape difference
with assert_raises(ValueError):
check_consistent_shape(X_train, y_train, y_train, y_test,
y_train, y_test)
def test_evaluate_print(self):
X_train, y_train, X_test, y_test = generate_data(
n_train=self.n_train,
n_test=self.n_test,
contamination=self.contamination)
evaluate_print('dummy', y_train, y_train * 0.1)
def test_evaluate_print(self):
X_train, y_train, X_test, y_test = generate_data(
n_train=self.n_train,
n_test=self.n_test,
contamination=self.contamination)
evaluate_print('dummy', y_train, y_train * 0.1)
import numpy as np
from sklearn.utils import shuffle
from matplotlib import pyplot as plt
from linear_regression.linear_regression import LinearRegressor
from utils.data import generate_data
x, y = generate_data()
x_shuffle, y_shuffle = shuffle(x, y, random_state=0)
lr = LinearRegressor(input_size=2)
lr.train(x_shuffle, y_shuffle, iter=10000, alpha=0.001)
plt.title("Loss per Iteration")
plt.xlabel("Iteration(s)")
plt.ylabel("Loss")
plt.plot(lr.historical_error)
plt.show()
t, s = lr.decision_boundry(start=-0.4, stop=1, step=0.01)
plt.title("Accuracy per Epoch")
plt.plot(lr.historical_accuracy)
plt.xlabel("Iteration(s)")
plt.ylabel("Accuracy")
plt.show()
plt.title('Classification Data')
plt.scatter(x[0:500, 0], x[0:500, 1], label='Class 1', marker='x')
plt.scatter(x[500:1000, 0], x[500:1000, 1], label='Class 2', marker='v')
plt.legend(['Class 1', 'Class 2'], loc='lower left')
plt.show()
