pint.templates.lctemplate.adaptive_samples

pint.templates.lctemplate.adaptive_samples(func, npt, log10_ens=3, nres=200)[source]

func should have a .cdf method.

NB – log10_ens needs to be a scalar!

The idea is to return a set of points on [0,1] which are approximately distributed uniformly in F(phi) and thus more densely sample the peaks. First, the cdf is evaluated on nres points. Then npt estimates of the inverse cdf are obtained by linear interpolation.