mosaicperm.statistics.quantile_maxcorr_stat¶
- mosaicperm.statistics.quantile_maxcorr_stat(residuals: array, qs: array | None = array([0.5, 0.6, 0.7, 0.8, 0.9, 0.95, 0.99]), **kwargs) array[source]¶
Quantiles of maximum (absolute) correlation statistic. This is defined as:
\[\begin{align*} \text{Quantile}\left( \left\{\max_{j \ne i} |\hat C_{i,j}|\right\}_{i=1}^p, q \right) \end{align*}\]for each q in
qs, where \(\hat C_{i,j}\) is the empirical correlation between the residuals of subjects i and j.- Parameters:
- residualsnp.array
(n_obs, n_subjects) array of residuals
- qsnp.array
Array of quantiles.
- kwargsdict
kwargs for
active_subsetpreprocessing function.`
- Returns:
- qmcnp.array
Vector of quantiles of maximum correlations.