Skip to content

1539: Fix numerical issues in prop_diff_uncond_exact - #1542

Open
danielinteractive wants to merge 7 commits into
mainfrom
1539-sas-discrepancy
Open

1539: Fix numerical issues in prop_diff_uncond_exact#1542
danielinteractive wants to merge 7 commits into
mainfrom
1539-sas-discrepancy

Conversation

@danielinteractive

@danielinteractive danielinteractive commented Oct 4, 2026

Copy link
Copy Markdown
Collaborator

closes #1539

@wwojciech

wwojciech commented Oct 4, 2026 •
edited
Loading

Copy link
Copy Markdown
Contributor

Thank you @danielinteractive , your idea looks great to me :) I will check this tomorrow morning in details (as I do not want to work on Sunday :) ).

@wwojciech wwojciech mentioned this pull request Oct 4, 2026
Closed
@github-actions

github-actions Bot commented Oct 4, 2026 •
edited
Loading

Copy link
Copy Markdown
Contributor

badge

Code Coverage Summary

Filename                                   Stmts    Miss  Cover    Missing
---------------------------------------  -------  ------  -------  ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
R/abnormal_by_baseline.R                     101       3  97.03%   242, 244-245
R/abnormal_by_marked.R                        88       8  90.91%   94-98, 281, 283-284
R/abnormal_by_worst_grade.R                   94       3  96.81%   215, 217-218
R/abnormal_lab_worsen_by_baseline.R          159      10  93.71%   205-208, 213, 215-216, 459-461
R/abnormal.R                                  78       2  97.44%   222, 224
R/analyze_variables.R                        320      11  96.56%   593-596, 818-823, 831
R/analyze_vars_in_cols.R                     178      14  92.13%   178, 221, 235-236, 238, 246-254
R/bland_altman.R                              92       1  98.91%   46
R/combination_function.R                       9       0  100.00%
R/compare_variables.R                         35       0  100.00%
R/control_incidence_rate.R                    10       0  100.00%
R/control_logistic.R                           7       0  100.00%
R/control_step.R                              23       1  95.65%   58
R/control_survival.R                          16       0  100.00%
R/count_cumulative.R                         115       4  96.52%   74, 270-271, 273
R/count_missed_doses.R                        89       4  95.51%   206-209
R/count_occurrences_by_grade.R               169       8  95.27%   178, 386, 388, 465, 467, 469, 473-474
R/count_occurrences.R                        137      10  92.70%   119, 262-264, 330-332, 334, 338-339
R/count_patients_events_in_cols.R             67       1  98.51%   60
R/count_patients_with_event.R                 73       2  97.26%   220, 223
R/count_patients_with_flags.R                 93       2  97.85%   234, 236
R/count_values.R                              61       2  96.72%   193, 196
R/cox_regression_inter.R                     154       0  100.00%
R/cox_regression.R                           161       0  100.00%
R/coxph.R                                    165       7  95.76%   190-194, 236, 251, 259, 265-266
R/d_pkparam.R                                406       0  100.00%
R/decorate_grob.R                            116       0  100.00%
R/desctools_binom_diff.R                     621      64  89.69%   53, 88-89, 125-126, 129, 199, 223-232, 264, 266, 286, 290, 294, 298, 353, 356, 359, 362, 422, 430, 439, 444-447, 454, 457, 466, 469, 516-517, 519-520, 522-523, 525-526, 593, 604-616, 620, 663, 676, 680
R/df_explicit_na.R                            45       0  100.00%
R/estimate_multinomial_rsp.R                  86       4  95.35%   65, 212, 214-215
R/estimate_proportion.R                      245       6  97.55%   88, 99, 255, 257-258, 389
R/fit_rsp_step.R                              36       0  100.00%
R/fit_survival_step.R                         36       0  100.00%
R/formatting_functions.R                     190       2  98.95%   141, 276
R/g_forest.R                                 636      47  92.61%   306, 309, 368, 370, 528, 615, 628, 632-633, 638-639, 652, 668, 715, 744, 819, 828, 834, 853, 908-928, 931, 942, 961, 1016, 1019, 1154-1159
R/g_ipp.R                                    133       0  100.00%
R/g_km.R                                     354      57  83.90%   285-288, 307-309, 363-366, 400, 428, 432-475, 482-486
R/g_lineplot.R                               261      22  91.57%   222, 397-404, 443-453, 562, 570
R/g_step.R                                    68       1  98.53%   108
R/g_waterfall.R                               47       0  100.00%
R/h_adsl_adlb_merge_using_worst_flag.R        73       0  100.00%
R/h_biomarkers_subgroups.R                    91      23  74.73%   40-42, 84-103
R/h_cox_regression.R                         110       0  100.00%
R/h_incidence_rate.R                          45       0  100.00%
R/h_km.R                                     510      39  92.35%   147, 199-204, 297, 388, 390-391, 402-404, 423, 430-431, 433-435, 443-445, 470, 475-478, 661-664, 1118-1127
R/h_logistic_regression.R                    468       3  99.36%   203-204, 273
R/h_map_for_count_abnormal.R                  54       0  100.00%
R/h_pkparam_sort.R                            15       0  100.00%
R/h_response_biomarkers_subgroups.R           77      12  84.42%   50-55, 107-112
R/h_response_subgroups.R                     178      18  89.89%   257-270, 329-334
R/h_stack_by_baskets.R                        64       1  98.44%   89
R/h_step.R                                   178       0  100.00%
R/h_survival_biomarkers_subgroups.R           73       6  91.78%   111-116
R/h_survival_duration_subgroups.R            207      18  91.30%   259-271, 336-341
R/imputation_rule.R                           17       0  100.00%
R/incidence_rate.R                           103       7  93.20%   68-73, 242
R/logistic_regression.R                      102       0  100.00%
R/missing_data.R                              26       5  80.77%   39, 62-63, 96, 106
R/odds_ratio.R                               157       4  97.45%   270-273
R/prop_diff_test.R                           234       2  99.15%   298, 300
R/prop_diff.R                                632       8  98.73%   386, 388, 868, 1034, 1142, 1740, 1744, 1747
R/prune_occurrences.R                         57       0  100.00%
R/response_biomarkers_subgroups.R            124      10  91.94%   88-91, 270-275
R/response_subgroups.R                       252      16  93.65%   100-105, 271-275, 280, 282-283, 310-311
R/riskdiff.R                                  65       4  93.85%   94-97
R/rtables_access.R                            38       0  100.00%
R/score_occurrences.R                         20       1  95.00%   124
R/split_cols_by_groups.R                      49       0  100.00%
R/stat.R                                      59       0  100.00%
R/summarize_ancova.R                         174       2  98.85%   355-356
R/summarize_change.R                          72       3  95.83%   175, 177-178
R/summarize_colvars.R                         13       1  92.31%   75
R/summarize_coxreg.R                         172       0  100.00%
R/summarize_glm_count.R                      269      10  96.28%   129-130, 202-203, 459-463, 596
R/summarize_num_patients.R                   121      10  91.74%   122-124, 244, 248, 252-253, 337-338, 340
R/summarize_patients_exposure_in_cols.R      155       7  95.48%   58, 232-233, 237, 357-358, 362
R/survival_biomarkers_subgroups.R            136      10  92.65%   117-122, 228-231
R/survival_coxph_pairwise.R                  154       9  94.16%   55-56, 124, 138, 145, 149, 288, 290-291
R/survival_duration_subgroups.R              250      15  94.00%   124-129, 268-273, 286, 288-289
R/survival_time.R                            128       1  99.22%   261
R/survival_timepoint.R                       153       2  98.69%   320, 322
R/utils_checkmate.R                           93       0  100.00%
R/utils_default_stats_formats_labels.R       201       0  100.00%
R/utils_factor.R                              87       1  98.85%   99
R/utils_ggplot.R                             112       0  100.00%
R/utils_grid.R                               126       5  96.03%   164, 279-286
R/utils_rtables.R                            125       9  92.80%   39, 46, 414-415, 537-541
R/utils_split_funs.R                          52       2  96.15%   82, 94
R/utils.R                                    161       7  95.65%   131, 134, 137, 141, 150-151, 345
TOTAL                                      12606     567  95.50%

Diff against main

Filename         Stmts    Miss  Cover
-------------  -------  ------  -------
R/prop_diff.R       +2      +1  -0.15%
TOTAL               +2      +1  -0.01%

Results for commit: 7f84131

Minimum allowed coverage is 80%

♻️ This comment has been updated with latest results

@github-actions

github-actions Bot commented Oct 4, 2026 •
edited
Loading

Copy link
Copy Markdown
Contributor

Unit Tests Summary

    1 files     89 suites   1m 37s ⏱️
1 032 tests 1 013 ✅  19 💤 0 ❌
2 652 runs  1 911 ✅ 741 💤 0 ❌

Results for commit 7f84131.

♻️ This comment has been updated with latest results.

@github-actions

github-actions Bot commented Oct 4, 2026 •
edited
Loading

Copy link
Copy Markdown
Contributor

Unit Test Performance Difference

Test Suite $Status$ Time on main $±Time$ $±Tests$ $±Skipped$ $±Failures$ $±Errors$
g_forest 💚 $23.42$ $-3.89$ $0$ $0$ $0$ $0$
Additional test case details
Test Suite $Status$ Time on main $±Time$ Test Case
prop_diff 👶 $+0.03$ prop_diff_uncond_exact_preserves_ties_in_the_SAS_regression_example
prop_diff 👶 $+0.14$ prop_diff_uncond_exact_respects_response_and_group_reversal_with_ties

Results for commit 639ac86

♻️ This comment has been updated with latest results.

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@wwojciech wwojciech Awaiting requested review from wwojciech

At least 1 approving review is required to merge this pull request.

Assignees

@danielinteractive danielinteractive

Labels

priority to refine

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

Discrepancy between SAS and prop_diff_uncond_exact()

2 participants

@danielinteractive @wwojciech