visual.kaito 0.1.0

First public release.

New functions

• boxplot3d() and boxplot3d_interactive(): triaxial (three continuous variables) box plots comparing groups, with four whisker conventions (Tukey, fixed-percentile, mean +/- SD, letter-value). The _interactive() version draws a single rotatable 3D plotly scene with live controls (whisker convention, raw/standardized scale, 2D axis-pair panels, color-by any categorical column); boxplot3d() draws the same four conventions as three linked 2D projections using base graphics only.
• boxplot3d_significance(): per-axis and joint (3D) significance testing to accompany the box plots above, combining a parametric test (Welch t-test/ANOVA per axis, MANOVA for the joint test) and a non-parametric test (Wilcoxon/Kruskal-Wallis per axis, a dependency-free permutation PERMANOVA for the joint test) in parallel.
• ttest_plot(): a didactic 2D plot of the t-distribution density with the critical (rejection) region shaded and the observed t-statistic marked, from raw data or from a manual t/df pair.
• ttest_plot3d(): two-group bivariate density comparison on two continuous variables, drawn as overlapping 3D density surfaces, with per-axis t-tests and a joint Hotelling’s T-squared test.
• manova_plot3d(): the >2-group generalization of ttest_plot3d(), with a one-way MANOVA omnibus test (Wilks’ lambda) and Bonferroni, Tukey, Fisher’s LSD, and Dunnett post-hoc comparisons.
• chisq_plot3d(): association between two categorical variables as a 3D bar chart of standardized (Pearson) residuals, with Cochran’s-rule and Simpson’s-paradox (descriptive + formal Gail-Simon) diagnostics.
• paired_plot3d(): repeated-measures/pre-post comparison on two continuous variables at once, with paired t-tests per variable and a joint paired Hotelling’s T-squared test.
• irt_plot3d(): Item Response Theory item calibration map (via mirt) drawing every item’s characteristic curve(s) in 3D against a shared ability (theta) axis, with a linked ability/difficulty histogram panel.

Shared interactive features

• Every plotly-based 3D function ships live, pre-computed controls (opacity, view/projection buttons, and function-specific selectors) that recompute instantly in the browser without calling R again.
• chisq_plot3d() and irt_plot3d() additionally offer a “3D / 2D” orthographic view toggle and a lang = c("es", "en") argument translating every on-plot label, hover string, and message.

Documentation

• An introductory vignette (vignette("visual-kaito")) covers all nine functions with runnable examples and a comparison to related packages (ggstatsplot, ggpubr, rstatix, effectsize, superb).
• A citation entry (citation("visual.kaito")) via CITATION.cff and inst/CITATION, including a Zenodo DOI.