Returns bootstrap-based stability for feature weights. For PLS, these
are bootstrap ratios (BSR = salience / SE). Other methods may use
different stability metrics.
Usage
stability(x, k = NULL, threshold = NULL, as_neurovol = FALSE, ...)
Arguments
- x
An mva_result or pls_result object
- k
Component index or vector (NULL = all)
- threshold
Optional threshold for masking
- as_neurovol
Convert to NeuroVol if mask available
- ...
Additional arguments
Value
Matrix of stability values, or NeuroVol