TL_GPRSM.utils package

Submodules

TL_GPRSM.utils.metrics module

TL_GPRSM.utils.metrics.mae(y_true, y_pred)

Mean absolute error

Parameters:

  • y_true (np.array) – true values

  • y_pred (np.array) – predicted values

Returns:

MAE

Return type:

float

TL_GPRSM.utils.metrics.mape(y_true, y_pred)

Mean absolute percentage error

Parameters:

  • y_true (np.array) – true values

  • y_pred (np.array) – predicted values

Returns:

MAPE

Return type:

float

TL_GPRSM.utils.metrics.me(y_true, y_pred)

Mean error

Parameters:

  • y_true (np.array) – true values

  • y_pred (np.array) – predicted values

Returns:

ME

Return type:

float

TL_GPRSM.utils.metrics.mpe(y_true, y_pred)

Mean percentage error

Parameters:

  • y_true (np.array) – true values

  • y_pred (np.array) – predicted values

Returns:

MPE

Return type:

float

TL_GPRSM.utils.metrics.mse(y_true, y_pred)

Mean squared error

Parameters:

  • y_true (np.array) – true values

  • y_pred (np.array) – predicted values

Returns:

MSE

Return type:

float

TL_GPRSM.utils.metrics.msle(y_true, y_pred)

Mean squared logarithmic error

Parameters:

  • y_true (np.array) – true values

  • y_pred (np.array) – predicted values

Returns:

MSLE

Return type:

float

TL_GPRSM.utils.metrics.r2_index(y_true, y_pred)

R2 index

Parameters:

  • y_true (np.array) – true values

  • y_pred (np.array) – predicted values

Returns:

R2 index

Return type:

float

TL_GPRSM.utils.metrics.rmse(y_true, y_pred)

Root mean squared error

Parameters:

  • y_true (np.array) – true values

  • y_pred (np.array) – predicted values

Returns:

RMSE

Return type:

float

TL_GPRSM.utils.metrics.rmsle(y_true, y_pred)

Root mean squared logarithmic error

Parameters:

  • y_true (np.array) – true values

  • y_pred (np.array) – predicted values

Returns:

RMSLE

Return type:

float

TL_GPRSM.utils.metrics.rmspe(y_true, y_pred)

Root mean squared percentage error

Parameters:

  • y_true (np.array) – true values

  • y_pred (np.array) – predicted values

Returns:

RMSPE

Return type:

float

TL_GPRSM.utils.metrics.wmae(y_true, y_pred, weights)

Weighted mean absolute error

Parameters:

  • y_true (np.array) – true values

  • y_pred (np.array) – predicted values

  • weights (np.array) – weights

Returns:

WMAE

Return type:

float

TL_GPRSM.utils.metrics.wmape(y_true, y_pred, weights)

Weighted mean absolute percentage error

Parameters:

  • y_true (np.array) – true values

  • y_pred (np.array) – predicted values

  • weights (np.array) – weights

Returns:

WMAPE

Return type:

float

TL_GPRSM.utils.sampling module

TL_GPRSM.utils.sampling.latin_hypercube_sampling(sampling_num, dim, center=False, seed=None)

Latin hypercube sampling

Parameters:

  • sampling_num (int) – number of sampling

  • dim (int) – dimension of sampling

  • center (bool, optional) – is center. Defaults to False.

  • seed (int, optional) – random seed. Defaults to None.

Returns:

sampling points (number of samples x number of features)

Return type:

numpy.ndarray

TL_GPRSM.utils.sampling.random_sampling(sampling_num, dim, seed=None)

Random sampling

Parameters:

  • sampling_num (int) – number of sampling

  • dim (int) – dimension of sampling

  • seed (int, optional) – random seed. Defaults to None.

Returns:

sampling points (number of samples x number of features)

Return type:

numpy.ndarray

TL_GPRSM.utils.sampling.uniform_scaling(sampling_points, scale_mins, scale_maxs)

Uniform scaling

Parameters:

  • sampling_points (np.array) – sampling points (number of samples x number of features)

  • scale_mins (list[float,...]) – min vals of each feature

  • scale_maxs (list[float,...]) – max vals of each feature

Returns:

scaled sampling points (number of samples x number of features)

Return type:

np.array

Module contents