regressioninc.testing.real module

Functions for generating and visualising real-valued testing data

regressioninc.testing.real.generate_linear(coef: ndarray, n_samples: int, intercept: float = 0)

Generate real-valued linear testing data

regressioninc.testing.real.add_gaussian_noise(data: ndarray, loc: float = 0, scale: float = 3) → ndarray

Add gaussian noise to an array

Parameters:

• data (np.ndarray) – The data to add the noise to

• loc (float, optional) – The location (mean) of the gaussian, by default 0. Usually this should# be left as 0.

• scale (float, optional) – The scale (or standard deviation) of the noise, by default 3

Returns:

Data with noise added

Return type:

np.ndarray

regressioninc.testing.real.add_outliers(y: ndarray, outlier_percent: float = 5, outlier_mult=3) → ndarray

Add outliers to a 1-D observations array

regressioninc.testing.real.linear_real_with_leverage()

regressioninc.testing.real.linear_real_with_outliers_and_leverage()

regressioninc.testing.real.plot_1d(X: ndarray, y: ndarray, coefs: Dict[str, Tuple[ndarray, float]] | None = None) → Figure

regressioninc.testing.real.get_X_plane(X: ndarray) → ndarray

regressioninc.testing.real.plot_2d(X: ndarray, y: ndarray, coefs: Dict[str, Tuple[ndarray, float]] | None = None) → Figure