The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.

News - bpca R package

1.5-0 (2026-05-25) - Faria, J. C.

Breaking changes

• var.rdf() now returns a structured list of class 'var.rdf' instead of a plain character data.frame. The three fields are: $display (character data.frame with '*', '', and '-' entries, for tabular display), $numeric (numeric matrix of absolute percentage differences), and $flagged (logical matrix, TRUE where the discrepancy exceeds limit). Code that accessed bp$var.rd as a data.frame directly must be updated to use bp$var.rd$display.
• summary.bpca() no longer accepts the presentation argument. The printing behaviour previously triggered by presentation = TRUE is now handled by the new print.summary.bpca() S3 method; calling print(summary(bp)) produces the same output.

New S3 methods

• print.var.rdf(): printing a var.rdf object (e.g. print(bp$var.rd)) automatically displays the character table ($display).
• print.summary.bpca(): separates computation from presentation following R’s S3 convention. summary.bpca() now only computes and returns the summary object; print.summary.bpca() handles console output.

Internal refactoring (no public API change)

• New file R/bpca-utils.R introduces four shared internal helpers that eliminate duplicated code across the package:
  • .center_scale(x, center, scale): centering and scaling logic previously duplicated between bpca.default() and dt.tools().
  • .cosine_matrix(m): pairwise cosine similarity between rows of a matrix, previously duplicated between var.rbf() and dt.tools().
  • .compute_var_factor(coobj, covar): automatic variable-scaling factor for biplots, previously duplicated between plot.bpca.2d() and plot.bpca.3d().
  • .pc_axis_labels(eigenvalues, dims): axis labels in "PC1 (38.5%)" format, previously duplicated between plot.bpca.2d() and plot.bpca.3d().
• plot.bpca.2d() decomposed from a single 634-line function into three files:
  • R/plot.bpca.2d.helpers.R: eight internal drawing functions (.draw_obj(), .draw_var(), .draw_var_seg(), .draw_circles(), .draw_axis_cross(), .proj_on_direction(), .solve_orthogonal_intersection(), .draw_circles_at()), extracted from closures that were recreated on every call.
  • R/plot.bpca.2d.types.R: ten internal type functions (.plot_type_bp() through .plot_type_rv()), one per switch branch, each independently readable and testable.
  • R/plot.bpca.2d.R: reduced to ~160 lines — validation, state preparation, context list (ctx) construction, and dispatch via switch().
• pc.names in bpca.default() is now generated once and reused for svdx.scal$v, g.scal, and hl.scal column names (previously generated twice).
• The fragile while loop in plot.bpca.2d(type = 'ww') — which relied on an NA sentinel at the end of chull() output — was replaced by a for loop with circular indexing over all convex-hull edges, including the closing edge.

Testing

• Test suite expanded from 3 to 45 tests. New coverage includes: all four factorization methods (hj, sqrt, jk, gh); all four centering modes (center = 0:3); all four internal helpers (.center_scale, .cosine_matrix, .compute_var_factor, .pc_axis_labels); var.rbf and var.rdf (including the new list structure); dt.tools; qbpca; summary.bpca and print.summary.bpca; all ten 2D plot types; 3D static plot; plot.qbpca; and bpca.prcomp.

1.4-3 (2026-05-11) - Faria, J. C.

Changes

• Cleaned core functions to improve readability for AI-assisted analysis.
• Manual improvements were made to some parts of the code that were refactored by AI.
• Refactored core computational functions for safer input handling and improved edge-case behavior (bpca.default, bpca.prcomp, dt.tools, qbpca, var.rbf, var.rdf).
• Refactored plotting internals (plot.bpca.2d, plot.bpca.3d, plot.qbpca) to reduce duplication and improve readability while preserving public API.
• Improved summary and table method internals (summary.bpca, xtable.bpca, print.xtable.bpca) and fixed minor string-handling issues.
• Enhanced 2D and 3D biplot diagnostics for improved visual clarity.
• Improved HTML table defaults in print.xtable.bpca for type = "html" while keeping LaTeX behavior unchanged.

Documentation

• Updated demo scripts across demo/ (2D/3D, static/dynamic, and utilities) to align examples with the refactored plotting code.
• Updated man/plot.Rd and man/qbpca.Rd to reflect the latest plotting behavior and usage.
• Expanded vignettes/bpca-overview.Rmd and added vignettes/latex-bpca.Rnw to support the updated guidance and examples.
• Documented print(x, type = "html") in man/print.xtable.Rd, the HTML overview vignette section, and README.md (use print() for S3 dispatch; results = 'asis' in R Markdown for HTML tables).
• Refined the qbpca plotting workflow to improve internal organization and keep graphical outputs consistent.
• print.xtable.bpca now supports type = "html" (via print(x, type = "html", ...)) for HTML table output, including column labels with eigenvalues; LaTeX layout is unchanged for the default type = "latex".
• Standardized and modernized package documentation across man/*.Rd.
• Updated package metadata files (DESCRIPTION, NAMESPACE, INDEX) for consistency with the current repository and dependency layout.
• Migrated the legacy NEWS file to NEWS.md for improved rendering on GitHub and pkgdown.
• Refactored vignettes/latex-bpca.Rnw for clearer didactic flow, improved visual layout, and updated examples.
• Updated vignettes/latex-bpca.pdf with the new vignette formatting.
• Simplified the HTML example in vignettes/bpca-overview.Rmd to use print(tbl, type = "html") directly.
• Regenerated vignettes/bpca-overview.html with the updated HTML table rendering.
• Removed prebuilt vignette artifacts from inst/doc/ to keep the repository clean.

Testing

• Added a new testthat suite under tests/testthat/ covering core bpca workflows, summary/plot smoke checks, and xtable output behavior for HTML and LaTeX.
• Updated package metadata to enable automated tests (Suggests: testthat (>= 3.0.0) and Config/testthat/edition: 3 in DESCRIPTION).

1.3-10 (2026-04-25) - Faria, J. C.

• Added support for vectorized var.pos in 2D and 3D plots, enabling precise manual label placement to reduce overlap.
• Fixed misaligned variable-to-object projections in plot.bpca.2d (type = "eo") by rewriting projection logic to properly account for scaling.

1.3-9 (2026-04-21) - Faria, J. C.

• Optimized plot.bpca.3d rendering performance when using rgl.
• Fixed an issue in plot.bpca.3d that prevented variable vectors from using distinct colors.
• Improved framing behavior in plot.bpca.2d and plot.bpca.3d, reducing the need for manual adjustments with var.factor, xlim, ylim, and zlim.
• Improved package and function documentation.

1.3-8 (2025-10-15) - Faria, J. C.

• Improved package and function documentation.

1.3-7 (2023-11-23) - Faria, J. C.

• Improved package and function documentation.

1.3-6 (2023-11-20) - Faria, J. C.

• Removed documentation notes as required by CRAN.

1.3-5 (2023-11-18) - Faria, J. C.

• Adjusted file encoding as required by CRAN.
• Performed substantial updates and improvements across package and function documentation.

1.3-4 (2021-03-23) - Faria, J. C.

• Completely redid the vignette to make it simpler, more direct, and more instructive.

1.3-3 (2021-03-19) - Allaman, I. B.

• Corrected a class-related bug in xtable.bpca.
• Fixed a bug in print.xtable.bpca; updated related commands and aux_com1 handling so sanitize.rownames.function works properly.

1.3-2 (2020-04-01) - Allaman, I. B.

• Created xtable.bpca and print.xtable.bpca.
• Deprecated and removed latex.bpca, print.latex.bpca, and summary.latex.bpca; corresponding .Rd files were removed.
• Updated NAMESPACE and DESCRIPTION accordingly.

1.3-1 (2018-06-16) - Allaman, I. B.

• Updated latex.bpca: changed footenotes default from "" to NULL.

1.3-0 (2018-06-07) - Allaman, I. B.

• Added stats, graphics, and grDevices imports in NAMESPACE to align with CRAN requirements.

1.2-2 (2013-11-23) - Faria, J. C.

• Added LaTeX table export for reduction summaries (latex.bpca).
• Improved parts of the source code.
• Added dataset marina.rda.
• Improved documentation.
• Removed unnecessary require calls for packages already attached via Depends (rgl and scatterplot3d).
• Released to CRAN.

1.2-1 (2012-12-12) - Faria, J. C.

• Added Ivan Bezerra Allaman as co-author in the project workflow.
• Added new plot options.
• Added new center options.
• Added new scale options.
• Added new summary options.
• Restricted release to testers.

1.2-0 (2012-10-06) - Faria, J. C.

• Improved parts of the source code.
• Removed var.position from the package; when needed, users can swap variable/object positions with bpca(t(dad)).
• Restricted release to testers.

1.0-10 (2012-02-20) - Faria, J. C.

• Applied cosmetic updates to plot.bpca.3d.

1.0-9 (2011-08-22) - Faria, J. C.

• Applied cosmetic updates to summary.

1.0-8 (2011-04-21) - Faria, J. C.

• Applied cosmetic updates to documentation.

1.0-7 (2011-04-20) - Faria, J. C.

• Improved summary method.
• Removed the comparison demo.

1.0-6 (2011-04-19) - Faria, J. C.

• Improved aliases for summary and plot methods in documentation.

1.0-5 (2011-04-09) - Faria, J. C.

• Fixed platform NOTE status in plot.bpca.3d (matrix(0, nc = 3) partial argument match of nc to ncol).

1.0-4 (2011-03-22) - Faria, J. C.

• Updated plot.bpca.2d and plot.bpca.3d to show, by default, the retained variation for each principal component on axis labels.
• Improved summary method.

1.0-3 (2009-06-10) - Faria, J. C.

• Updated plot.bpca.2d and plot.bpca.3d to accept xlim, ylim, zlim, xlab, ylab, and zlab.
• Reworked demos.
• Included PDF documentation in the package.

1.0-2 (2008-07-15) - Faria, J. C.

• Released to CRAN.
• Applied English corrections (by Clarice).
• Removed demo scripts using obj.identify = TRUE.

1.0-1 (2008-07-01) - Faria, J. C.

• First restricted release (testers only).

These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.