Add `partition_loo()` method from {sperrorest}

[pat-s]

Alternative to spatial_buffering from {blockCV}

library(sperrorest)

parti <- partition_loo(ecuador, buffer = 500)

parti[[1]] <- parti[[1]][1:3]
class(parti) <- "represampling"

plot(parti, ecuador)

^{Created on 2021-04-10 by the reprex package (v1.0.0)}

[pat-s]

This should be fully equal to rsmp("spcv_buffer") as a circular buffer is drawn around the observation representing the test set.

Hence, I am not gonna add it as it would be a fully redundant method.