htfa.tfa.standardize_data¶

htfa.tfa.standardize_data(X, mode)[source]¶

Z-score data (V, T) before fitting.

Parameters:

Xarray_like, shape (V, T): Data, one row per voxel and one column per sample.
mode{‘auto’, True, False}: 'auto' z-scores each row (voxel) across columns when T > 1 and leaves single-sample data unchanged. True z-scores each row when T > 1 and z-scores across all voxels when T == 1. False leaves the data unchanged. Constant rows (or a constant single column under True) become 0.

Returns:

ndarray, shape (V, T): The standardized data as float64.

Raises:

ValueError: If mode is not 'auto', True or False.

Examples

>>> import numpy as np
>>> from htfa.tfa import standardize_data
>>> standardize_data(np.array([[1.0, 3.0], [2.0, 2.0]]), "auto")
array([[-1.,  1.],
       [ 0.,  0.]])
>>> standardize_data(np.array([[1.0], [3.0]]), True)
array([[-1.],
       [ 1.]])