You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
ggdotplotstats() now labels the y axis with the name of the y
variable when ylab = NULL (default), as documented. Previously, no y
axis title was shown unless ylab was supplied.
ggcorrmat() now uses all numeric variables when cor.vars = NULL is
passed explicitly, as documented, instead of failing with an error.
ggcorrmat() now shows a single sample size in the legend whenever all
correlation pairs share the same sample size. Previously, missing values in
any column, including non-numeric columns not used in the correlations,
switched the legend to minimum, mode, and maximum sample sizes.
grouped_gghistostats() now computes its default common bin width from
the non-missing values of x, as gghistostats() does. Previously, missing
values were counted in the sample size, making the bins too narrow.
ggbarstats() no longer displays empty proportion test labels above the
bars when x has only one level, matching ggpiestats().
gghistostats() and grouped_gghistostats() now draw the histogram when all
values of x are identical (including a single observation). Previously, the
default bin width was zero, so no bins were drawn.
ggbarstats() and ggpiestats() now show the test of independence in the
subtitle when x or y has an empty level, such as an unused factor level.
Previously, the subtitle was silently dropped. This fix comes from {statsExpressions}.
gghistostats() and ggdotplotstats() with a one-sided alternative now
report a one-sided confidence interval for the effect size in the subtitle,
matching the one-sided p-value. This fix comes from {statsExpressions}.