One workflow for the three analyses: the analysis, interpret() for its axes, ggfacto() for its graph, hierarchical_clust() for its clusters. See the guide: https://bricenocenti.github.io/ggfacto/articles/ggfacto.html.
multiple_correspondence_analysis(), principal_component_analysis(), correspondence_analysis() (from a tabxplor::tab()): the analyses, weights and variables selected as in tab().interpret(): the table of an analysis' axes, with its eigenvalues below; eigenvalues() alone.ggfacto(): the graph of any of the three analyses; ggpca(), the individuals of a PCA and its biplot.hierarchical_clust(): HCPC()'s clusters, in linear memory, written into the data with mutate(); clust_tab() describes them.axis_coord(), is_in_analysis(): an analysis' coordinates, and its population, back into the data. name_axes() names the axes on graphs and tables.excl = takes exact level names; missing answers become <VAR>.NA levels, excluded by default (specific MCA).filter = analyses a subset; the whole data frame is then passed to every later function.clust =, names =, interactive =, lang =; options(ggfacto.widget_dir) writes widgets to their own file in documents.profiles = TRUE and active_tables = "active" are the defaults: the crosstables of the Burt table are in the tooltips.MCA2() keeps the 0.3.2 fit on individuals.HCPC_tab() (use clust_tab()), the cah and dat arguments (use clust and data), and %>% are soft-deprecated.pers_or_plot() is removed; ggi(iframe =, pixel_width =) are ignored.Frequency line of a tooltip could exceed 100 %.y/n disappeared from the graphs and tables.tabxplor:: version 1.2.1HCPC_tab() was not working well with numeric variables.HCPC_tab()ggmca_initial_dims()ggmca_with_base_ref()ggmca_3d()PCA2()pca_interpret()ggpca_cor_circle()ggpca_3d()mean_sd_tab()ggmca() can now properly color profiles of answer with cah (HCPC clusters) levels.reverse_axes in ggmca() to invert left and right or up and down.mca_interpret was not working with MCA(excl = )ggmca, profiles of answers were plain wrong (because of the unanticipated results of a change in code)