DoE Analysis: Factorial and Interaction plots by JTPetter · Pull Request #470 · jasp-stats/jaspQualityControl

JTPetter · 2026-05-05T14:44:48Z

Adds factorial and interaction plots to DoE analysis
Verified unit tests for the new plots
Fixes [Feature Request]: Add factorial plots in DOE jasp-issues#3889

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

codecov-commenter · 2026-05-29T15:04:23Z

Codecov Report

❌ Patch coverage is 82.50000% with 56 lines in your changes missing coverage. Please review.
✅ Project coverage is 84.89%. Comparing base (8c33634) to head (a6f27ea).

Files with missing lines	Patch %	Lines
R/doeAnalysis.R	82.50%	56 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #470      +/-   ##
==========================================
- Coverage   84.96%   84.89%   -0.08%     
==========================================
  Files          18       18              
  Lines       10012    10332     +320     
==========================================
+ Hits         8507     8771     +264     
- Misses       1505     1561      +56

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

Copilot

Pull request overview

Copilot reviewed 3 out of 37 changed files in this pull request and generated 1 comment.

vandenman · 2026-09-18T14:28:02Z

+      stop(gettextf("Error computing emmeans for terms '%s': %s. Prediction fallback failed: %s.",
+                     paste(termNames, collapse = ", "), conditionMessage(e), conditionMessage(e2)))


This comment should be addressed @JTPetter

JohnnyDoorn · 2026-07-03T11:48:46Z

build bot build

github-actions · 2026-07-03T12:01:16Z

✅ Build Complete!

Your JASP module bundles have been successfully built for macOS (ARM) and Windows.

👉 Click here to download the artifacts

(Scroll to the bottom of the page to find the "Artifacts" section).

JohnnyDoorn

the placeholder plot shows when loading the analysis, which shouldn't be the case
the options could be laid out across 2 columns
For the interaction plots, would it maybe increase the clarity of the figures if you work with faded colors for the confidence bands, similar to the other descriptives plots in JASP? They currently do look nice and funky, but it's not so easy to get a clear idea of the interactions.
the plots appear in the top, changing the plots triggers the main results which is probably not what is needed.
to plot the interaction for continuous predictors there is some stratifying going on that can create messy labels - maybe you can take a look at how it's done in regression for stratifying continuous predictors: jasp-stats/jaspRegression#514

JTPetter · 2026-07-10T13:56:30Z

Thanks for the review! Regarding your first point, are you referring to the plot below? This placeholder is not related to the new plots and has always been there. EJ asked for some empty plot to be there as a place holder and not start an analysis completely empty.

JTPetter · 2026-07-10T15:30:17Z

@JohnnyDoorn I addressed all other comments. I did not understand the last point though. Do you maybe have an example where the stratifying creates weird labels?

JTPetter · 2026-09-04T14:30:30Z

@JohnnyDoorn any news on this?

vandenman

Some basic remarks about the code, still have to run this in JASP though.

vandenman · 2026-09-18T14:28:02Z

+      stop(gettextf("Error computing emmeans for terms '%s': %s. Prediction fallback failed: %s.",
+                     paste(termNames, collapse = ", "), conditionMessage(e), conditionMessage(e2)))


This comment should be addressed @JTPetter

vandenman · 2026-09-18T14:30:36Z

+  if (includeCi) {
+    pred <- stats::predict(model, newdata = grid, se.fit = TRUE)
+    df <- stats::df.residual(model)
+    tcrit <- stats::qt((1 + level) / 2, df)
+    emmean <- pred$fit
+    se <- pred$se.fit
+    lower <- emmean - tcrit * se
+    upper <- emmean + tcrit * se
+  } else {
+    emmean <- stats::predict(model, newdata = grid)
+  }
+
+  out <- grid[, termNames, drop = FALSE]
+  out$emmean <- emmean
+  if (includeCi) {
+    out$lower.CL <- lower
+    out$upper.CL <- upper
+  }
+
+  return(out)


Perhaps put out <- grid[, termNames, drop = FALSE] first to avoid doing twice if (includeCi) {?

Suggested change

if (includeCi) {

pred <- stats::predict(model, newdata = grid, se.fit = TRUE)

df <- stats::df.residual(model)

tcrit <- stats::qt((1 + level) / 2, df)

emmean <- pred$fit

se <- pred$se.fit

lower <- emmean - tcrit * se

upper <- emmean + tcrit * se

} else {

emmean <- stats::predict(model, newdata = grid)

}

out <- grid[, termNames, drop = FALSE]

out$emmean <- emmean

if (includeCi) {

out$lower.CL <- lower

out$upper.CL <- upper

}

return(out)

out <- grid[, termNames, drop = FALSE]

if (includeCi) {

pred <- stats::predict(model, newdata = grid, se.fit = TRUE)

df <- stats::df.residual(model)

tcrit <- stats::qt((1 + level) / 2, df)

emmean <- pred$fit

se <- pred$se.fit

out$lower <- emmean - tcrit * se

out$upper <- emmean + tcrit * se

} else {

emmean <- stats::predict(model, newdata = grid)

}

out$emmean <- emmean

return(out)

vandenman · 2026-09-18T14:31:28Z

+  return(out)
+}
+
+.doeAnalysisNormalizeCiLevel <- function(level) {


Is this necessary? It looks like an AI-ism (😅) of something that should always be enforced by QML rather than by R?

vandenman · 2026-09-18T14:33:08Z

+    factorType <- .doeAnalysisPlotFactorType(factorName, options)
+    if (factorType == "continuous") {
+      emm[["x"]] <- as.numeric(emm[[factorName]])
+      xBreaks <- sort(unique(c(min(emm[["x"]], na.rm = TRUE), max(emm[["x"]], na.rm = TRUE))))


Suggested change

xBreaks <- sort(unique(c(min(emm[["x"]], na.rm = TRUE), max(emm[["x"]], na.rm = TRUE))))

xBreaks <- sort(unique(c(range(emm[["x"]], na.rm = TRUE))))

range(x) returns c(min(x), max(x))

vandenman · 2026-09-18T14:34:57Z

+
+    if (xFactorType == "continuous") {
+      emm[["x"]] <- as.numeric(emm[[factorA]])
+      xBreaks <- sort(unique(c(min(emm[["x"]], na.rm = TRUE), max(emm[["x"]], na.rm = TRUE))))


range. Also the code around lines 2241 looks like a duplicate of the code here. Perhaps extract a common helper?

EJWagenmakers · 2026-09-20T18:33:45Z

FYI: this is key functionality :-)

JTPetter requested a review from Copilot May 5, 2026 14:45

Copilot started reviewing on behalf of JTPetter May 5, 2026 14:46 View session

Copilot AI reviewed May 5, 2026

JTPetter force-pushed the doeEmPlot branch from 4522144 to db4791f Compare May 14, 2026 16:11

JTPetter force-pushed the doeEmPlot branch from db4791f to b52ee95 Compare May 29, 2026 14:37

JTPetter force-pushed the doeEmPlot branch from b52ee95 to a6f27ea Compare June 2, 2026 14:44

JTPetter requested a review from Copilot June 5, 2026 13:53

Copilot started reviewing on behalf of JTPetter June 5, 2026 13:53 View session

JTPetter requested a review from JohnnyDoorn June 5, 2026 13:54

Copilot AI reviewed Jun 5, 2026

View reviewed changes

JohnnyDoorn requested changes Jul 3, 2026

View reviewed changes

JTPetter added 4 commits July 9, 2026 16:49

DoE emmeans plots sketch

55d8c27

factorial plots v2 and unit tests for what works already

5a0de3e

Fix with cont. factors and squared terms and update unit tests

75be3b8

Extend axes, adjust tests, new tests

da357e8

JTPetter force-pushed the doeEmPlot branch from a6f27ea to da357e8 Compare July 9, 2026 14:54

fallback snaps

35ffbcf

adress feedback

47e7d72

JTPetter requested a review from JohnnyDoorn July 10, 2026 15:30

vandenman requested changes Sep 18, 2026

View reviewed changes

		stop(gettextf("Error computing emmeans for terms '%s': %s. Prediction fallback failed: %s.",
		paste(termNames, collapse = ", "), conditionMessage(e), conditionMessage(e2)))

	xBreaks <- sort(unique(c(min(emm[["x"]], na.rm = TRUE), max(emm[["x"]], na.rm = TRUE))))
	xBreaks <- sort(unique(c(range(emm[["x"]], na.rm = TRUE))))

Conversation

JTPetter commented May 5, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Uh oh!

codecov-commenter commented May 29, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

vandenman Sep 18, 2026

Choose a reason for hiding this comment

Uh oh!

JohnnyDoorn commented Jul 3, 2026

Uh oh!

github-actions Bot commented Jul 3, 2026

✅ Build Complete!

Uh oh!

JohnnyDoorn left a comment

Choose a reason for hiding this comment

Uh oh!

JTPetter commented Jul 10, 2026

Uh oh!

JTPetter commented Jul 10, 2026

Uh oh!

JTPetter commented Sep 4, 2026

Uh oh!

vandenman left a comment

Choose a reason for hiding this comment

Uh oh!

vandenman Sep 18, 2026

Choose a reason for hiding this comment

Uh oh!

vandenman Sep 18, 2026

Choose a reason for hiding this comment

Uh oh!

vandenman Sep 18, 2026

Choose a reason for hiding this comment

Uh oh!

vandenman Sep 18, 2026

Choose a reason for hiding this comment

Uh oh!

vandenman Sep 18, 2026

Choose a reason for hiding this comment

Uh oh!

EJWagenmakers commented Sep 20, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

JTPetter commented May 5, 2026 •

edited

Loading

codecov-commenter commented May 29, 2026 •

edited

Loading