`skcriteria.core.stats` module

Stats helper for the DecisionMatrix object.

class skcriteria.core.stats.DecisionMatrixStatsAccessor(dm)[source]

Calculate basic statistics of the decision matrix.

Kind of statistic to produce:

‘corr’ : Compute pairwise correlation of columns, excluding NA/null values.
‘cov’ : Compute pairwise covariance of columns, excluding NA/null values.
‘describe’ : Generate descriptive statistics.
‘kurtosis’ : Return unbiased kurtosis over requested axis.
‘mad’ : Return the mean absolute deviation of the values over the requested axis.
‘max’ : Return the maximum of the values over the requested axis.
‘mean’ : Return the mean of the values over the requested axis.
‘median’ : Return the median of the values over the requested axis.
‘min’ : Return the minimum of the values over the requested axis.
‘pct_change’ : Percentage change between the current and a prior element.
‘quantile’ : Return values at the given quantile over requested axis.
‘sem’ : Return unbiased standard error of the mean over requested axis.
‘skew’ : Return unbiased skew over requested axis.
‘std’ : Return sample standard deviation over requested axis.
‘var’ : Return unbiased variance over requested axis.

corr(method='kendall', **kwargs)[source]

Compute pairwise correlation of criteria columns.

By default the Kendall rank correlation coefficient (tau) is used, which is appropriate for ordinal or non-normally distributed data.

Parameters:

method (str or callable, default "kendall") – Method of correlation. Accepted values are "pearson", "kendall", "spearman", or a callable with signature (Series, Series) -> float. See pandas.DataFrame.corr() for details.
kwargs – Other keyword arguments are passed to the underlying pandas.DataFrame.corr() method.

Returns:

Symmetric DataFrame of shape (n_criteria, n_criteria) with the pairwise correlation coefficients between criteria. Diagonal entries are 1.0.

Return type:

pd.DataFrame

skcriteria.core.stats module