Version 1.10.5 for CRAN

DESCRIPTION

@@ -1,8 +1,8 @@
 Package: emmeans
 Type: Package
 Title: Estimated Marginal Means, aka Least-Squares Means
-Version: 1.10.4.900003
-Date: 2024-10-02
+Version: 1.10.5
+Date: 2024-10-14
 Authors@R: c(person("Russell V.", "Lenth", role = c("aut", "cre", "cph"), 
     email = "russell-lenth@uiowa.edu"),
     person("Balazs", "Banfai", role = "ctb"),

NEWS.md

@@ -7,6 +7,7 @@ title: "NEWS for the emmeans package"
   * Fix for inconsistent contrasts in case of missing levels (#508, #509)
   * Fix for using nuisance variables with proportional weights (#510)
   * New function `with_emm_options()` to run code with options temporarily set
+  * Tweak to optimal-digits output that shows `SE` to 3 significant digits
 
 
 

R/summary.R

@@ -1360,6 +1360,8 @@ print.summary_emm = function(x, ..., digits=NULL, quote=FALSE, right=TRUE, expor
 
     est = x[[estn]]
     if (get_emm_option("opt.digits") && is.null(digits)) {
+        if (!is.null(x$SE))
+            x$SE = signif(x$SE, 3)
         if (!is.null(x[["SE"]]))
             tmp = est + x[["SE"]] * cbind(rep(-2, nrow(x)), 0, 2)
         else if (!is.null(x[["lower.HPD"]]))

README.md

@@ -90,6 +90,13 @@ They can always be found [for the latest CRAN version](https://cran.r-project.or
 For the latest release notes on this development version, see the 
 [NEWS file](https://github.com/rvlenth/emmeans/blob/master/NEWS.md)
 
+## Rounding
+For its summary output, **emmeans** uses an optimal-digits algorithm that rounds
+results to about the number of digits that are useful, relative to estimates' 
+confidence limits. This avoids cluttering the output, but it is unlike other R
+results, which are typically less round. If this is annoying to you, there is an 
+option (`opt.digits = FALSE`) that disables the optimal-digits routine.
+
 ## "Tidiness" can be dangerous
 I see more and more users who are in a terrible hurry to get results. They develop
 a "workflow" where they plan-out several steps at once and pipe them

vignettes/FAQs.Rmd

@@ -39,6 +39,8 @@ This vignette contains answers to questions received from users or posted on dis
  12. [I asked for a Tukey adjustments, but that's not what I got](#notukey)
  12. [`emmeans()` completely ignores my P-value adjustments](#noadjust)
  13. [`emmeans()` gives me pooled *t* tests, but I expected Welch's *t*](#nowelch)
+ 14. [I want to see more digits](#rounding)
+
 
 [Index of all vignette topics](vignette-topics.html)
 
@@ -419,5 +421,12 @@ you will obtain incorrect results. Get the model right *first*.
 
 [Back to Contents](#contents)
 
+## I want to see more digits {#rounding}
+The summary display of `emmeans()` and other functions uses an optimal-digits
+routine that shows only relevant digits. If you temporarily want to see more
+digits, add `|> as.data.frame()` to the code line. In addition, piping
+`|> data.frame()` will disable formatting of test statistics and P values (but 
+unfortunately will also suppress messages below the tabular output).
+
 [Index of all vignette topics](vignette-topics.html)