Changelog • ggfortify

Fixed failed tests in test-stat and test-ts with ggplot2 4.0.0.

Fixed failed tests with ggplot2 4.0.0.

Fixed an issue with updated n.censor from survfitms output when using survival v3.6.1.

Resolved warnings in R CMD check related to S3 generic/method consistency.

Removed autoplot() support for objects in dlm library.
Resolved warnings in R CMD check.

Switched to use range of leverages to check for constant leverage to be consistent with plot.lm().
Switched to use MIT license.

Added support for survfit objects with multiple stratification variables.
Fixed bug in rendering frames for clustering.

Used QuickStartExample to access x and y explicitly to fix CRAN error.
Fixed the warning related to the non-uniform sampler used in tests.
Added various missing visible bindings for global variables.

Added support of Silhouette plot for cluster::silhouette objects, thanks to @damirpolat.
Fixed issue with autoplot.lm(which = 2) when residuals have AsIs type, thanks to @richierocks.
Added markdown to Suggests. For context, please see this issue.

This is a minor release of the package with fixes in tests to be compatible with survival (>=3.1-12).

This is a minor release of the package with fixes for CRAN check results.

This is a minor release of the package with fixes to support new major release of survival (>=3.1-7).

This is a minor release of the package with a couple of fixes for CRAN check results.

Fixed CRAN check results for R-devel regarding changes in the default method for generating from a discrete uniform distribution used in sample().

Updated tests to work with ggplot2 v2.3.0.

Added label.show.legend argument to hide/show the legend of text labels.
ggmultiplot arranges a single ggplot instance with layout that can now be specified by the user.
Enhanced unit tests: 1) CRAN MKL build can pass; tests are skipped if the required packages are not installed; several test checks are more robust to upstream package changes.

Fixed incorrect y-axis label for plotting cv.glmnet objects.
Removed incorrect y-axis label with scales::percent for ggdistribution.
Added support of autoplot.lm() for character variables, thanks to @lselzer.
Fixed tests for forecast::ets() to be compatible with newer version of forecast package, thanks to @robjhyndman.
Added support for arbitrary functions in autoplot.survfit(), thanks to @jonathon-love.

Removed gglagplot in favor of forecast::gglagplot.
Added support of survival::survfitms, thanks to @yoursdearboy.
Temporarily dropped support for xts and timeSeries objects. We will try fix them in the next release.

Survival plot now displays legend as it is (without prefix / reordering), thanks to @DarioS.

Support ROCR::performance, thanks to @austin3dickey
PCA allows to choose the components to plot, thanks to @lselzer.
LM autoplot fails when a formula contains polynomial regression.
GLM autoplot fails when a family function is binomial.

Our white paper is on R Journal. Type citation("ggfortify") for more info.
Support splines::basis.
Support several raster ojects, such as RasterBrick, RasterCommon, RasterLayer, and RasterStack.
ggmultiplot objects can now be saved correctly via ggplot2::ggsave().

Components of prcomp and princomp are now scaled. Scaling can be disabled by specifying scale = 0.
autoplot.t now supports stacked option.
Support maps and sp packages.
Support ggrepel.
Support ggbiplot.
More compatible with recent version of forecast, dplyr, tidyr, etc.

Compatible with ggplot2 v2.0.0.
Support glmnet.
Support multiplots from a list of classes supported by autoplot.
Support ggmultiplot extraction getter / setter.
Support ggmultiplot arithmetics.

First release on CRAN.