Subsetting psis objects

[jgabry]

The PR introduces a subset() method for "psis" objects. This is useful if you want to drop some of the contents of a "psis" object without breaking it. For example, bayesplot has an option for this for ppc_loo_intervals() (https://mc-stan.org/bayesplot/reference/PPC-loo.html?q=subset#arguments) implemented via an internal function but it would be good if loo provided this functionality and bayesplot can call it.

[codecov-io]

Codecov Report

Merging #110 into master will decrease coverage by 2.76%.
The diff coverage is 91.5%.

@@            Coverage Diff             @@
##           master     #110      +/-   ##
==========================================
- Coverage   97.73%   94.97%   -2.77%     
==========================================
  Files          16       19       +3     
  Lines        1281     2048     +767     
==========================================
+ Hits         1252     1945     +693     
- Misses         29      103      +74

Impacted Files	Coverage Δ
R/loo_compare.R	94.79% <100%> (+0.6%)	⬆️
R/waic.R	100% <100%> (ø)	⬆️
R/psis_approximate_posterior.R	71.42% <59.37%> (-24.58%)	⬇️
R/loo.R	91.87% <70.73%> (-5.63%)	⬇️
R/loo_compare.psis_loo_ss_list.R	84.82% <84.82%> (ø)
R/loo_approximate_posterior.R	92.1% <92.1%> (ø)
R/print.R	96.96% <94.44%> (-1.15%)	⬇️
R/loo_subsample.R	95.4% <95.4%> (ø)
R/effective_sample_sizes.R	97.27% <0%> (-0.23%)	⬇️
... and 14 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 44a9312...e9c900b. Read the comment docs.

[avehtari]

@jgabry is this still work in progress?