pypomp.maths.logmeanexp_se

pypomp.maths.logmeanexp_se(x: Any, axis: None = None, ignore_nan: bool = False) → float[source]
pypomp.maths.logmeanexp_se(x: Any, axis: int, ignore_nan: bool = False) → ndarray

A jack-knife standard error for the log-likelihood estimate calculated via logmeanexp(). For comparison with R-pomp::logmeanexp, note that np.std divides by n whereas R-sd divides by (n-1), so np.var gives the Gaussian MLE and R-var gives the unbiased estimator.

Parameters:

  • x (array-like) – collection of log-likelihoods

  • axis (int or None) – axis along which to compute the SE. If None (default), compute over the entire array.

  • ignore_nan (bool) – if True, drop NaNs before computing.