This version (0.3.2) adds as new features scales and statistics that help with the creation of volcano and quadrant plots, such as used with transcriptomics and metabolomics data. A few rough edges remaining in the features added in versions 0.3.0. and 0.3.1 have been polished out. Two issues raised in Bitbucket about the documentation, highlighted some incomplete explanations. These explanations have now been expanded. One important change to the documentation of statistics whose returned values may change depending on arguments is the addition of an example of the use of geom_debug() from package ‘gginnards’ showing how to print to the R console the data returned by statistics, which is the input received by the paired geometries. The User Guide needs still some work, scheduled for the next release. Package documentation is available at https://docs.r4photobiology.info/ggpmisc/ as a web site.
Plots created using the new statistics and scales are shown below. In the quadrant plot, which observations were labelled and highlighted was decided automatically based on local 2D density. Counts for each quadrant are computed on the fly. As the plot is non-the-less created using the grammar of graphics, little if any of the flexibility of ‘ggplot2’ is lost.
NOTE: The new version of ‘ggpmisc’ is on its way to CRAN.
This post is not to announce something related to my own packages, but to highlight Rob Hyndman’s new packages for working with time-series data using a tidy approach. If you have to deal with time series in R, you should have a look at these packages and read the posts with examples of their use at the Hyndsight blog.
I suspect more is to come soon, but for the time being have a look at what Rob Hyndman wrote yesterday and today in his blog.