Reference /
binference.diagnostics.importance_ess
binference.diagnostics.importance_ess
-
binference.diagnostics.importance_ess(weights)[source]
Kish’s effective sample size of importance weights,
\((\sum_s w_s)^2 / \sum_s w_s^2\) (normalised or not).
- Parameters:
- weightsarray_like (n,)
Importance weights.
