thin: Perform Stein thinning
In wilson-ye-chen/stein.thinning: Stein Thinning

Optimally select m points from n > m samples generated from a target distribution of d dimensions.

1	thin(smp, scr, m, stnd = TRUE, pre = "id")

`smp`	n-by-d matrix of n d-dimensional sample points.
`scr`	n-by-d matrix of gradients of log target at `smp`.
`stnd`	optional logical, either TRUE (default) or FALSE, indicating whether or not to standardise the colums of `smp`.
`pre`	optional string, either "id" (default), "med", "sclmed", or "smpcov", specifying the preconditioner to be used. Alternatively, a numeric string can be passed as the single length-scale parameter of an isotropic kernel.