def test_linear_regressor_predict_not_fitted():
"""Test of `LinearRegressor` class with prediction without fit."""
lr = LinearRegressor(learning_rate=0.1, iterations=1, standardize=False)
X = np.array([1])
with pytest.raises(NotFitted):
lr.predict(X)
def test_linear_regressor_score_not_fitted():
"""Test of `LinearRegressor` class with score calculation without fit."""
lr = LinearRegressor(learning_rate=0.1, iterations=1, standardize=False)
X = np.array([1])
y = np.array([1])
with pytest.raises(NotFitted):
lr.score(X, y)
def test_linear_regressor_dataset_inconsistancy():
"""Test of `LinearRegressor` with dataset inconsistancy."""
lr = LinearRegressor(learning_rate=0.1, iterations=1)
X = np.array([[1], [1]])
y = np.array([[1]])
with pytest.raises(InvalidInput):
lr.fit(X, y)
def test_linear_regressor_normal():
"""Test of `normal` method of `LinearRegressor` class."""
lr = LinearRegressor(learning_rate=0.1, iterations=1, standardize=False)
X = np.array([1])
y = np.array([1])
lr.normal(X, y)
assert lr.score(X, y) == 1
def test_linear_regressor_standardized():
"""Test of `LinearRegressor` class with standardization."""
lr = LinearRegressor(learning_rate=0.1, iterations=1, standardize=True)
X = np.array([1])
y = np.array([1])
lr.fit(X, y)
assert lr.score(X, y) == 1
def test_linear_regressor_history_disabled():
"""Test of `LinearRegressor` when history disabled."""
lr = LinearRegressor(learning_rate=0.1, iterations=1, history=False)
assert lr.history is None
X = np.array([1])
y = np.array([1])
lr.fit(X, y)
assert lr.history is None
def test_linear_regressor_history_enabled():
"""Test of `LinearRegressor` when history enabled."""
lr = LinearRegressor(learning_rate=0.1, iterations=1, history=True)
assert lr.history == []
X = np.array([1])
y = np.array([1])
lr.fit(X, y)
assert lr.history is not None
# Create the input features and target data frame
df = pd.DataFrame(houses.data, columns=houses.feature_names)
df["target"] = pd.Series(houses.target)
# Select the input feature
X = df["RM"]
X = X.values.reshape(X.shape[0], 1)
# Select the target
y = df["target"]
y = y.values.reshape(y.shape[0], 1)
# Train the model
lr = LinearRegressor(optimizer=MBGD(learning_rate=0.05,
iterations=70,
history=True),
standardize=True)
lr.fit(X, y)
# Plot the results
figure, (a0, a1) = plt.subplots(1,
2,
num="Linear Regression",
figsize=(16, 9),
gridspec_kw={"width_ratios": [3, 1]})
figure.suptitle("Linear Regression on Boston house prices dataset",
fontsize=16)
a0.plot(X, y, "b.")
a0.plot(
range(3, 10),