mescla.emma.rank.rank_summary

mescla.emma.rank.rank_summary(eigenvalues, Z=None, threshold=0.9)[source]

Scree table with every retention criterion side by side.

Parameters:

  • eigenvalues (ndarray)

  • Z (ndarray, optional) – Standardised data. Required for the parallel-analysis column.

  • threshold (float, default 0.9) – Cumulative-variance target.

Returns:

DataFrame – One row per component: eigenvalue, proportion and cumulative proportion of variance, the broken-stick expectation, and a tick for each rule that would retain that component. The recommended k from each rule is recorded in .attrs["recommendations"].

Return type:

pd.DataFrame