add extra tests

[malcolmbarrett]

To support the changes in #45

[Copilot]

Pull Request Overview

This PR adds comprehensive test coverage for numeric vs. psw weight equivalence and backward compatibility to support changes in #45. The tests ensure that balance computation functions work identically with both numeric weights and psw (propensity score weights) objects, maintaining backward compatibility while introducing new functionality.

Key changes:

• Added extensive test suite for numeric/psw weight equivalence across all balance functions
• Created tests for the extract_weight_data() utility function
• Included edge cases, performance comparisons, and mixed weight type scenarios

Reviewed Changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
tests/testthat/test-numeric-psw-equivalence.R	Comprehensive test suite covering backward compatibility, equivalence testing, edge cases, and performance comparisons between numeric and psw weights
tests/testthat/test-extract_weight_data.R	Test coverage for utility function that extracts numeric data from weight objects, handling various input types and edge cases

Comments suppressed due to low confidence (2)

tests/testthat/test-numeric-psw-equivalence.R:45

• [nitpick] This test verifies identical results but doesn't test what happens when results should differ. Consider adding a test case where different estimand types might produce different behaviors or edge cases where equivalence might break down.

test_that("numeric and psw weights produce identical results", {

tests/testthat/test-numeric-psw-equivalence.R:300

• Performance timing tests can be flaky and unreliable, especially on CI systems with variable load. Consider using a larger tolerance or focusing on functional correctness rather than performance timing.

  expect_true(time_psw[["elapsed"]] < time_numeric[["elapsed"]] * 1.5)