Update regress.R

[AmeerD]

Fix for Cox model with missing data (see this issue: https://github.com/statdivlab/rigr/issues/148). The following changes were made:

• Subsetted the weights variable after missing cases are removed.
• Removed the check on the length of the subsetting variable after subsetting as it is not used further down.

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 66.66667% with 1 lines in your changes are missing coverage. Please review.

Project coverage is 72.23%. Comparing base (a4489c8) to head (d6e8ec8).

:exclamation: Current head d6e8ec8 differs from pull request most recent head 1a9bf66. Consider uploading reports for the commit 1a9bf66 to get more accurate results

Files	Patch %	Lines
R/regress.R	66.66%	1 Missing :warning:

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #151 +/- ## ======================================= Coverage 72.23% 72.23% ======================================= Files 23 23 Lines 3738 3739 +1 ======================================= + Hits 2700 2701 +1 Misses 1038 1038 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[svteichman]

Overall this is looking good. There are a few additional updates that I would recommend:

• update the version from 1.0.5 to 1.0.6, to show that a patch has been added to address a bug
• add a test in the file "tests/testthat/test_regress.R" that either uses the example from the issue #151 or a simpler example of the bug and shows that the bug is now fixed (for this example, it seems sufficient to show that the code runs in this setting)

[svteichman]

This looks great! I'd recommend removing the .Rhistory file that is being added, and then it should be all set to merge! You should be able to do this by adding .Rhistory to the .gitignore file on your local fork, although it may require deleting the file in your remote github fork.

[svteichman]

Looks ready to merge to me!