Stats¶

best.stats.combine_gauss_distributions(mu1, std1, N1, mu2, std2, N2)¶: Recalculates a normal 1-D distribution given two subsets of data.

best.stats.combine_mvgauss_distributions(mu1, var1, N1, mu2, var2, N2)¶: Recalculates a normal n-D distribution given two subsets of data.

best.stats.compare_datasets(dataset, states=['AWAKE', 'N2', 'N3', 'REM'])¶

Compares dataset consistency using KDE and prec-recall curves

Parameters: dataset (dict) – dict where the key is name of dataset for comparison; each dataset is represented by dict with 2 variables X, Y
Returns: dataset
Return type: dict

best.stats.get_class_count(Y, classes=None)¶: Returns a number of class appearance in the labels

best.stats.kl_divergence(mu1, std1, mu2, std2)¶

Parametric KL-Divergence between 2 normal 1-D distributions.

best.stats.kl_divergence_mv(mu1, var1, mu2, var2)¶

Multidimensional parametric KL-Divergence between 2 normal distributions.

best.stats.kl_divergence_nonparametric(pk, qk)¶: Calculates non-parametric KL-Divergence between two 1-D distributions given by 2 histograms with same bins.