Add weight variable and missing data handling options to rwa() by Copilot · Pull Request #22 · martinctc/rwa

Copilot · 2026-01-21T10:02:48Z

Summary

This branch adds support for weighted analysis and configurable missing data handling in RWA.

Changes

The changes made in this PR are:

Added use parameter to rwa() and rwa_multiregress() to control missing data handling in correlations (defaults to "pairwise.complete.obs" for backward compatibility)
Added weight parameter to rwa() and rwa_multiregress() for weighted RWA using cov.wt() (defaults to NULL)
Updated rwa() wrapper function to pass use and weight parameters through to rwa_multiregress()
Updated bootstrap functions to handle both parameters consistently
Added validation: weight must be numeric and positive; use must be valid cor() option
Added warning when weight/use parameters are used with logistic regression (they only apply to multiple regression)
Added comprehensive test coverage for both features
Refactored implementation to work with the new rwa() wrapper structure from master branch

Usage:

# Pairwise vs listwise deletion
rwa(df, "outcome", c("x1", "x2"), use = "complete.obs")  # listwise
rwa(df, "outcome", c("x1", "x2"), use = "pairwise.complete.obs")  # pairwise (default)

# Weighted analysis
df$survey_weight <- c(1.2, 0.8, 1.5, ...)
rwa(df, "outcome", c("x1", "x2"), weight = "survey_weight")

Check

roxygen2::roxygenise() has been run prior to merging to ensure that .Rd and NAMESPACE files are up to date.
The R CMD checks pass (at least one or more of devtools::check() or rhub::check_for_cran())

Change Type

Please check the type of change your PR introduces:

Bugfix
Feature (incl. changes to visualizations)
Technical design
Build related changes (e.g. tests)
Refactoring (no functional changes, no api changes)
Code style update (formatting, renaming) or Documentation content changes
Other (please describe):

(OPTIONAL) Note

Backward compatible - existing code continues to work unchanged. Both features integrate seamlessly with bootstrap confidence intervals and work only with multiple regression (not logistic regression). Implementation updated to work with the refactored rwa() wrapper structure from master branch.

Original prompt

This section details on the original issue you should resolve

<issue_title>weights and missings</issue_title>
<issue_description>Hi Martin, love the package, but I have 2 questions:

Is it possible to run the analysis with a separate weight variable?

If I'm correct the analysis is standard with listwise deletion of missings. is it also possible to run it with pairwise deletion?
Thanks!</issue_description>

<agent_instructions>Please use the given pull request template when creating a pull request out of this issue. </agent_instructions>

Comments on the Issue (you are @copilot in this section)
@martinctc Hi @SietseDost - currently we don't have any capability in `rwa` to support a separate weight variable. @stonid - do you by any chance have an implementation of this available, so we can look at including this in the package?
With your second question @SietseDost, this is also not available but I can include this as a feature in #6 or #7. The plan is to export the parameter for use in cor() here, in rwa():
  cor_matrix <-
    cor(thedata, use = "pairwise.complete.obs") %>%
    as.data.frame(stringsAsFactors = FALSE, row.names = NULL) %>%
    remove_all_na_cols() %>%
    tidyr::drop_na()
```</body></comment_new>
<comment_new><author>@stonid</author><body>
@SietseDost @martinctc Unfortunately, I don't have an implementation for weighting. The missing data option should be relatively easy to implement as a new feature as Martin noted.</body></comment_new>
</comments>

Fixes weights and missings #12

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.

Co-authored-by: martinctc <17925865+martinctc@users.noreply.github.com>

… consistent Co-authored-by: martinctc <17925865+martinctc@users.noreply.github.com>

Co-authored-by: martinctc <17925865+martinctc@users.noreply.github.com>

martinctc · 2026-01-22T12:20:03Z

@copilot - these new changes have significant conflicts with the new master branch. The implementation / fix for #12 has to be updated to work with the new refactored rwa(). Can you please update?

Co-authored-by: martinctc <17925865+martinctc@users.noreply.github.com>

Copilot · 2026-01-22T12:25:40Z

@copilot - these new changes have significant conflicts with the new master branch. The implementation / fix for #12 has to be updated to work with the new refactored rwa(). Can you please update?

Updated in commit af35bb4. The implementation has been refactored to work with the new rwa() wrapper structure:

Added use and weight parameters to both rwa() and rwa_multiregress()
The wrapper now passes these parameters through to rwa_multiregress()
Updated all tests and documentation to work with the new structure
Both features work only with multiple regression (warning added for logistic regression)
Bootstrap functionality updated to support both parameters

The implementation maintains full backward compatibility and resolves the conflicts with the master branch.

…com/martinctc/rwa into copilot/add-weight-variable-support

…te tests

…nctions

Copilot

Pull request overview

This PR adds two new parameters to the rwa() function to address user requests for weighted analysis and configurable missing data handling. The implementation adds a use parameter to control how missing data is handled when computing correlations (defaulting to "pairwise.complete.obs" for backward compatibility) and a weight parameter to enable weighted RWA using survey or importance weights. Both features are integrated with the existing bootstrap confidence interval functionality and work exclusively with multiple regression (not logistic regression).

Changes:

Added use and weight parameters to rwa() and rwa_multiregress() functions with comprehensive validation
Updated bootstrap functions to propagate both parameters through the analysis pipeline
Added warning when weight/use parameters are used with logistic regression
Added 217 lines of comprehensive test coverage for both features and their interactions

Reviewed changes

Copilot reviewed 8 out of 8 changed files in this pull request and generated 8 comments.

Show a summary per file

File	Description
R/rwa.R	Added parameters, validation logic, and warning for logistic regression; passes parameters to sub-functions
R/rwa_multiregress.R	Implemented core weighted correlation using cov.wt() and configurable missing data handling
R/bootstrap_rwa.R	Updated all bootstrap functions to handle use/weight parameters consistently
man/rwa.Rd	Documented new use and weight parameters for main wrapper function
man/rwa_multiregress.Rd	Documented new use and weight parameters with examples
tests/testthat/test-rwa.R	Added comprehensive test coverage for validation, edge cases, and parameter combinations
NEWS.md	Documented new features and improvements
NAMESPACE	Added imports for complete.cases and cov.wt

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

R/bootstrap_rwa.R

NEWS.md

man/rwa.Rd

man/rwa_multiregress.Rd

tests/testthat/test-rwa.R

R/rwa_multiregress.R

Initial plan

a11cde8

Copilot AI assigned Copilot and martinctc Jan 21, 2026

Copilot started work on behalf of martinctc January 21, 2026 10:03 View session

Copilot AI and others added 7 commits January 21, 2026 10:06

Implement missing data handling with use parameter

3ae8a0a

Co-authored-by: martinctc <17925865+martinctc@users.noreply.github.com>

Implement weight variable support for weighted RWA

e6bf523

Co-authored-by: martinctc <17925865+martinctc@users.noreply.github.com>

Update documentation and NEWS for new features

224662a

Co-authored-by: martinctc <17925865+martinctc@users.noreply.github.com>

Fix redundant as.data.frame conversion in bootstrap_rwa.R

52bf4e8

Co-authored-by: martinctc <17925865+martinctc@users.noreply.github.com>

Address code review feedback - fix error message and NA weight handling

abd9f38

Co-authored-by: martinctc <17925865+martinctc@users.noreply.github.com>

Fix weight validation to require positive values and make NA handling…

f7d7979

… consistent Co-authored-by: martinctc <17925865+martinctc@users.noreply.github.com>

Add trailing comma to function parameter list

be1a438

Co-authored-by: martinctc <17925865+martinctc@users.noreply.github.com>

Copilot AI changed the title ~~[WIP] Add support for separate weight variable in analysis~~ Add weight variable and missing data handling options to rwa() Jan 21, 2026

Copilot AI requested a review from martinctc January 21, 2026 10:17

Copilot finished work on behalf of martinctc January 21, 2026 10:17

Copilot started work on behalf of martinctc January 22, 2026 12:20 View session

Refactor to work with new rwa() wrapper and rwa_multiregress() structure

af35bb4

Co-authored-by: martinctc <17925865+martinctc@users.noreply.github.com>

Copilot finished work on behalf of martinctc January 22, 2026 12:26

martinctc added 4 commits January 22, 2026 13:00

Merge branch 'master' into copilot/add-weight-variable-support

05db37e

Merge branch 'copilot/add-weight-variable-support' of https://github.…

3d4bb29

…com/martinctc/rwa into copilot/add-weight-variable-support

feat: add support for weight variable in bootstrap functions and upda…

744eae2

…te tests

feat: add weight variable support and missing data handling to rwa fu…

a6c025b

…nctions

martinctc requested a review from Copilot January 23, 2026 15:14

Copilot started reviewing on behalf of martinctc January 23, 2026 15:15 View session

Copilot AI reviewed Jan 23, 2026

View reviewed changes

martinctc added 4 commits February 12, 2026 11:10

refactor: simplify redundant NA weight handling branches

adf13ad

docs: fix weight validation wording from non-negative to positive

4a561a3

docs: document that use param is ignored when weight is specified

0814262

docs: clarify misleading pairwise deletion test comment

400040b

martinctc added 2 commits February 12, 2026 11:12

docs: clarify n_used semantics for pairwise deletion

04f80ae

chore: document Rd

433d2d4

martinctc marked this pull request as ready for review February 27, 2026 15:37

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add weight variable and missing data handling options to rwa()#22

Add weight variable and missing data handling options to rwa()#22
Copilot wants to merge 19 commits intomasterfrom
copilot/add-weight-variable-support

Copilot AI commented Jan 21, 2026 •

edited

Loading

Uh oh!

martinctc commented Jan 22, 2026

Uh oh!

Copilot AI commented Jan 22, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

Copilot AI commented Jan 21, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Changes

Check

Change Type

(OPTIONAL) Note

Comments on the Issue (you are @copilot in this section)

Uh oh!

martinctc commented Jan 22, 2026

Uh oh!

Copilot AI commented Jan 22, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Copilot AI commented Jan 21, 2026 •

edited

Loading