Add pressure testing functions by pratikunterwegs · Pull Request #14 · vimc/vimcheck

pratikunterwegs · 2026-04-17T10:55:42Z

This PR adds functions from the pressure testing report.

Added impact diagnostics functions in R/fn_impact_diagnostics.R.
Added plotting preparation functions and plotting functions in R/fn_plotting_prep_impact_diagnostics.R and R/fn_plotting_impact_diagnostics.R.
Added dependencies diffdf and here.
Added vignettes on pressure testing functions and package design.
Minor change renaming files and reorganising documentation to be clearer.

Getting started

The easiest way to get started with reviewing is to check out the develop branch and render the package website using {pkgdown}: pkgdown::build_site(). View the vignette on pressure testing under Articles which explains the added functions.

Open task

Figure out what data can be added to pkg to allow for testing
Test plotting preparation functions
Test plotting functions
Add vignette documentation with example data
Rename file re: impact diagnostics to pressure testing? (?)

zegibney · 2026-04-17T15:49:51Z

+#' @export
+gen_combined_df <- function(prev_dat, df2, interest_cols, key_cols) {
+ # TODO: input checks
+ # TODO: df2 needs a better name


df2 is essentially a cleaned version of of the touchstone_new latest report on packit where vaccine names are cleaned, subregions appended etc. happy to take suggestions on a better way to do it

Thanks - could you suggest an informative name for the argument that somebody new to the package and the workflow would understand? data_current? data_cleaned? If you could pop a couple of lines in this thread about any expectations around it, I can add those to the function docs.

df_clean is fine, thanks. df_clean is the imported dependency report (currently named touchstone_new) with the following cleaning steps applied:
Malaria vaccine names amended
Subregions appended
Fixes for missing country names
Removal of now-redundant columns
It may even be sufficient to just apply these steps to df itself, I think I just kept them separate when developing the report so I didn't have to keep pulling a clean dependency when testing things out.

github-actions · 2026-04-22T15:00:35Z

This pull request:

Adds 4 new dependencies (direct and indirect)
Adds 1 new system dependencies
Removes 0 existing dependencies (direct and indirect)
Removes 0 existing system dependencies

Reach out on slack (#code-review or #help channels) to double check if there are base R alternatives to the new dependencies.

(Note that results may be inaccurate if you branched from an outdated version of the target branch.)

zegibney · 2026-04-23T14:02:07Z

+ # for scales formatting
+ .x <- NULL
+
+ # TODO: should NA-producing values (< 1) be removed?


I think previously we have used a lowerbound <- 1e-8 to filter out very small values

zegibney · 2026-04-23T14:06:24Z

+) {
+ checkmate::assert_data_frame(df, min.cols = 1L, min.rows = 1L)
+
+ # TODO: can we find checks for prev_data size in reln to df? rows? cols?


I'm not sure I fully understand? But prev_dat and df would likely be matched on columns

zegibney · 2026-04-23T14:14:56Z

+ variable <- rlang::arg_match(variable)
+ checkmate::assert_character(group_cols, min.len = 1L, any.missing = FALSE)
+
+ # TODO: check what a sensible upper limit might be


I think in my pressure testing report, an arbitrary limt was difference exceeding ± 100 times the interquartile range (IQR) for the relevant country, vaccine and activity_type.

zegibney · 2026-04-28T10:12:49Z

To discuss on call whether to keep pressure testing functions and rapid-model-run-impact function separate or combine for ease

github-actions · 2026-06-05T14:20:33Z

This pull request:

Adds 4 new dependencies (direct and indirect)
Adds 1 new system dependencies
Removes 0 existing dependencies (direct and indirect)
Removes 0 existing system dependencies

Reach out on slack (#code-review or #help channels) to double check if there are base R alternatives to the new dependencies.

(Note that results may be inaccurate if you branched from an outdated version of the target branch.)

github-actions · 2026-06-09T09:26:58Z

This pull request:

Adds 4 new dependencies (direct and indirect)
Adds 1 new system dependencies
Removes 0 existing dependencies (direct and indirect)
Removes 0 existing system dependencies

Reach out on slack (#code-review or #help channels) to double check if there are base R alternatives to the new dependencies.

(Note that results may be inaccurate if you branched from an outdated version of the target branch.)

github-actions · 2026-06-09T15:20:02Z

This pull request:

Adds 4 new dependencies (direct and indirect)
Adds 1 new system dependencies
Removes 0 existing dependencies (direct and indirect)
Removes 0 existing system dependencies

Reach out on slack (#code-review or #help channels) to double check if there are base R alternatives to the new dependencies.

(Note that results may be inaccurate if you branched from an outdated version of the target branch.)

github-actions · 2026-06-09T15:32:57Z

This pull request:

Adds 4 new dependencies (direct and indirect)
Adds 1 new system dependencies
Removes 0 existing dependencies (direct and indirect)
Removes 0 existing system dependencies

Reach out on slack (#code-review or #help channels) to double check if there are base R alternatives to the new dependencies.

(Note that results may be inaccurate if you branched from an outdated version of the target branch.)

xiangli313 · 2026-06-18T09:54:42Z

+ )
+ }
+
+ # data may have a `burden_outcome` column which should not be counted as


I think burden_outcome should be moved to the default key_cols COLNAMES_KEY_PRESSURE_TEST; and be picked up in the pressure_testing.Rmd L98-99.

I looked into adding this to COLNAMES_KEY_PRESSURE_TEST, and this causes the problem that these names are used in functions where the burden is expected to be in wide format, so a separate column for each burden type. I think the parsimonious course of action is to add this to the checked columns here, but to leave it out from the constant vector for now. See example function filter_invalid_trajectories() below this one, where the outcomes are expected as "*_averted".

I see. Thanks for the clarification.

xiangli313 · 2026-06-18T09:55:55Z

+flag_duplicates(eg_impact) %>%
+ select(
+ modelling_group, country, disease,
+ vaccine, activity_type, year, n_key


as mentioned ealier, burden_outcome may also be picked up here.

Added burden_outcome here.

xiangli313 · 2026-06-18T10:02:18Z

+ names_from = "burden_outcome",
+ values_from = "impact"
+)
+prev_df$support_type <- "other" # unsure what values this can take


We are moving from support_type ("total", "gavi", "other") to percentage of gavi support (numeric between 0-1). And this will not be part of burden processing or impact calculation. It will only appear in our out-facing impact extrapolation reports for funders; and highly dependent on funders incremental updates (twice a year). I think we can ignore this classifier for the pressure testing.

github-actions · 2026-06-18T14:16:35Z

This pull request:

Adds 4 new dependencies (direct and indirect)
Adds 1 new system dependencies
Removes 0 existing dependencies (direct and indirect)
Removes 0 existing system dependencies

Reach out on slack (#code-review or #help channels) to double check if there are base R alternatives to the new dependencies.

(Note that results may be inaccurate if you branched from an outdated version of the target branch.)

pratikunterwegs added 5 commits March 12, 2026 13:46

Add initial fns from pressure testing report
c8b4bc1

Move pressure testing plotting
db03465

Clean up funs and add constants
50f6b0d

Clean up fns and add constants
13d6244

Rename files
8aa8494

zegibney reviewed Apr 17, 2026
View reviewed changes

Comment threadR/fn_pressure_testing.R Outdated

zegibney reviewed Apr 17, 2026
View reviewed changes

Comment threadR/constants.R Outdated

zegibney reviewed Apr 17, 2026
View reviewed changes

Comment threadR/fn_pressure_testing.R Outdated

zegibney reviewed Apr 17, 2026
View reviewed changes

Comment threadR/fn_pressure_testing.R Outdated

pratikunterwegs added 11 commits April 20, 2026 16:55

Add checks and docs for pressure testing fns
de1eda6

Add pkg constants
46d3d97

Add pkg helpers
63e5871

WIP separate plot prep from plot code
ffff840

WIP add docs and pkg infra
d0237fb

Reorganise docs
4a52fbb

WIP plotting and prep fns impact diagnostics
bba30c5

Add constants
a455c5a

Update plot prep and plot fns
beb2a07

Fixes from initial test run
0424f17

Update infra, bump to v0.0.4
386be9b

Rename fn file, fix arg name
43753fb

zegibney reviewed Apr 23, 2026
View reviewed changes

zegibney reviewed Apr 28, 2026
View reviewed changes

pratikunterwegs self-assigned this Apr 29, 2026

pratikunterwegs changed the title ~~WIP adding pressure testing~~Add pressure testing functionsApr 29, 2026

pratikunterwegs mentioned this pull request Apr 30, 2026
Adding functions to produce impact, coverage and fvp plots #15
Closed

pratikunterwegs added 2 commits June 4, 2026 17:23

Fn filter dups only flags instead
474fdc3

More sensible checks on fns
4873b45

pratikunterwegs added 4 commits June 5, 2026 15:17

Add test data from vimpact
50b5389

Minor fixes to fns
5024eb9

Add test suite for impact fns
991d479

Update docs and NEWS
9663b5c

pratikunterwegs added the enhancement New feature or request label Jun 5, 2026

pratikunterwegs added 4 commits June 8, 2026 17:14

Minor edits to fns
72dff25

Add tests for plotting fns
97585b1

Update wordlist and add design vignette
113b551

WIP pressure testing vignette
d87f5a0

Minor updates, update vignette
1f1ac93

Fix docs issues
812311d

pratikunterwegs marked this pull request as ready for review June 9, 2026 15:36

xiangli313 reviewed Jun 18, 2026
View reviewed changes

Update vignette
8d566cd

xiangli313 approved these changes Jun 19, 2026
View reviewed changes

pratikunterwegs merged commit df208a4 into mainJun 19, 2026
11 checks passed

pratikunterwegs deleted the develop branch June 19, 2026 10:05

Uh oh!

Conversation

pratikunterwegs commented Apr 17, 2026• edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Getting started

Open task

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

github-actionsBot commented Apr 22, 2026

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

github-actionsBot commented Jun 5, 2026

Uh oh!

github-actionsBot commented Jun 9, 2026

Uh oh!

github-actionsBot commented Jun 9, 2026

Uh oh!

github-actionsBot commented Jun 9, 2026

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

github-actionsBot commented Jun 18, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

pratikunterwegs commented Apr 17, 2026•
edited
Loading