Overload matrix multiplication operator for rvars in R >= 4.3

[mjskay]

Summary

As promised on bluesky (hah), since in R >= 4.3 you can now implement %*% for S3 classes, this PR allows you to use %*% instead of %**% for matrix multiplying rvars.

There are also some minor maintenance things in there: it bumps to a dev version, adds that section in news, and re-documents.

Copyright and Licensing

By submitting this pull request, the copyright holder is agreeing to
license the submitted work under the following licenses:

• Code: BSD 3-clause (https://opensource.org/licenses/BSD-3-Clause)
• Documentation: CC-BY 4.0 (https://creativecommons.org/licenses/by/4.0/)

[codecov-commenter]

Codecov Report

Attention: 3 lines in your changes are missing coverage. Please review.

Comparison is base (c6ab463) 95.60% compared to head (7bc98df) 95.52%.

❗ Current head 7bc98df differs from pull request most recent head 263a5e6. Consider uploading reports for the commit 263a5e6 to get more accurate results

Files	Patch %	Lines
R/zzz.R	0.00%	2 Missing ⚠️
R/rvar-math.R	66.66%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #320      +/-   ##
==========================================
- Coverage   95.60%   95.52%   -0.08%     
==========================================
  Files          47       47              
  Lines        3686     3691       +5     
==========================================
+ Hits         3524     3526       +2     
- Misses        162      165       +3     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[mjskay]

The one missing line in code coverage probably can't be tested until they add more generics to matrixOps.

The check failures I am fixing now.

[github-actions]

This is how benchmark results would change (along with a 95% confidence interval in relative change) if bc26835 is merged into master:

•   :ballot_box_with_check:as_draws_array: 101ms -> 101ms [-0.99%, +0.47%]
•   :rocket:as_draws_df: 35.5ms -> 32.4ms [-10.27%, -7.38%]
•   :ballot_box_with_check:as_draws_list: 190ms -> 189ms [-2.33%, +0.73%]
•   :ballot_box_with_check:as_draws_matrix: 30.2ms -> 30.3ms [-1.7%, +2.52%]
•   :ballot_box_with_check:as_draws_rvars: 160ms -> 162ms [-0.38%, +3.03%]
•   :ballot_box_with_check:summarise_draws_100_variables: 721ms -> 718ms [-1.13%, +0.42%]
•   :ballot_box_with_check:summarise_draws_10_variables: 79.2ms -> 79.5ms [-0.37%, +1.04%]
  Further explanation regarding interpretation and methodology can be found in the documentation.

[github-actions]

This is how benchmark results would change (along with a 95% confidence interval in relative change) if 7bc98df is merged into master:

•   :ballot_box_with_check:as_draws_array: 99.7ms -> 101ms [-0.01%, +2.28%]
•   :rocket:as_draws_df: 34.5ms -> 31.5ms [-10.35%, -7.34%]
•   :ballot_box_with_check:as_draws_list: 180ms -> 183ms [-1.32%, +5.03%]
•   :ballot_box_with_check:as_draws_matrix: 29ms -> 29.2ms [-3.89%, +5.45%]
•   :ballot_box_with_check:as_draws_rvars: 155ms -> 154ms [-2.03%, +0.22%]
•   :ballot_box_with_check:summarise_draws_100_variables: 706ms -> 706ms [-0.52%, +0.36%]
•   :ballot_box_with_check:summarise_draws_10_variables: 78.4ms -> 78.2ms [-0.45%, +0.13%]
  Further explanation regarding interpretation and methodology can be found in the documentation.

[mjskay]

This should be ready now; the test failure in R 3.6 looks to be unrelated.

[paul-buerkner]

Looks good. Thank you!