0.1.9

• lm2() – (1) Mean column: Percentages (%) are shown only for factor levels and factor×factor interactions; numeric predictors and numeric×factor interactions show the plain mean (no %). A fallback computes factor-level percentages from the model frame when the model matrix is missing or column names do not match (e.g. with estimatr::lm_robust). (2) Baseline row: The baseline category (e.g. group (A) [baseline]) now shows the % of observations in the mean column, like the other factor levels.
• desc_var() – When the variable does not exist (e.g. desc_var(x) with no x), only the error is shown; the uninformative warning "restarting interrupted promise evaluation" is no longer emitted.
• scatter.gam() – (1) Accepts both direct and formula syntax: scatter.gam(x, y), scatter.gam(y ~ x, data = df), and scatter.gam(df$y ~ df$x) all work. (2) User-supplied ylim (and other plot args in ...) no longer triggers "formal argument matched by multiple actual arguments". (3) With formula syntax like mtcars$mpg ~ mtcars$hp, axis labels and title show only the variable name (e.g. xlab = "hp") instead of the full expression. (4) Lower (frequency) panel gets a 10% buffer on the y-axis so the top bar is not flush with the top of the plot.
• Message colors – All console messages that used red4 now use red2 for better visibility in dark-mode terminals (e.g. plot_freq(), plot_density(), plot_cdf() dropped-observation messages, clear(), twolines(), and exit() default).
• Speedup t.test2() – Performance improvements.
• clear() – Now works with a one-time permission authorization: on first use you are prompted to type "yes" to allow clearing the global environment, console, and plot; that choice is saved for future sessions. If you do not type "yes", you are asked again next time.
• plot_freq(), plot_density(), plot_cdf() – (1) Allow comparing two vectors (in addition to formula syntax). (2) Legend placed in the top 25% of the plot.
• table2() – Now supports comparison operators in variable names. When using expressions like table2(df$var>10), the dimension name displays the full expression (e.g., "var>10") instead of not displaying the var name at all. Long expressions (>30 characters) are automatically truncated with "..." for readability.