Efficiency improvements in predict

[palday]

Did behavior change? Did you add need features? If so, please update NEWS.md

• add entry in NEWS.md
• after opening this PR, add a reference and run docs/NEWS-update.jl to update the cross-references.

Should we release your changes right away? If so, bump the version:

• I've bumped the version appropriately

There are a couple of improvements here:

• using sortperm directly instead of faking it with sort(::Range)
• views
• skip actually computing the BLUPs for the hidden new model. They're nonsensical, the computation can fail because the new model hasn't actually been fit, and they'll be overwritten by the relevant entries from the old model.

There will be a merge conflict with #603 for NEWS.md, but I'll sort that out after the first of this is merged.

[codecov]

Codecov Report

Merging #604 (98ba01e) into main (3d868bb) will increase coverage by 0.03%.
The diff coverage is 100.00%.

❗ Current head 98ba01e differs from pull request most recent head 358ac1b. Consider uploading reports for the commit 358ac1b to get more accurate results

@@            Coverage Diff             @@
##             main     #604      +/-   ##
==========================================
+ Coverage   96.36%   96.39%   +0.03%     
==========================================
  Files          28       28              
  Lines        2695     2694       -1     
==========================================
  Hits         2597     2597              
+ Misses         98       97       -1     

Impacted Files	Coverage Δ
src/predict.jl	98.30% <100.00%> (-0.03%)	⬇️
src/linearmixedmodel.jl	98.24% <0.00%> (+0.17%)	⬆️

---

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 7f2b621...358ac1b. Read the comment docs.