This non-code-breaking update brings a few new features and tracks deprecations in package ‘tidyr’.
This is a major and code-breaking update. Naming conventions have been changed and several new spectra have been added. Previously included spectra have in most cases been recomputed and may slightly differ from earlier versions of the same data. In part this was done to reduce the size of the data objects, making it possible to include more spectra while keeping the size of the package reasonable. The number of spectra is more than 100, including several of the newest high CRI LEDs as well as some recent types for horticulture.
A handbook on how to do calculations used in photobiological research with R has been under preparation for a long time. An 80% complete version has been available through LeanPub for several years. The book describes the use of our suite of R packages and also of functions from base R and a few other packages available through CRAN. I (Pedro J. Aphalo) am the lead author, while Andreas Albert, T. Matthew Robson and Titta Kotilainen have contributed text and examples and feedback.
Although we haven’t had time to finish writing the book, I have been checking that the code examples work with current versions of R and packages. Rather recently I made small edits to a few code chunks that had stopped working and uploaded the new version of the PDF file to LeanPub.
Some minutes ago I tried to build all examples with R 4.2.o, a major update to R released some days ago. R 4.2.o contains a bug, that unluckily prevents a couple some examples in the book from running under MS-Windows. The already available patched version of R 4.2.0 solves this problem. Anyway, this bug affects only a very specific use related to data acquisition.
The book is sold for the amount the buyer wishes to pay, including getting it for free.
For Windows users the most significant changes are that now the binary distribution contains only a 64 bit executable. A new build chain is used and Rtools 4.2 is required to build from sources packages that include C, C++ or FORTRAN code. As is always the case, all CRAN packages are tested on CRAN itself.
Relevant to all operating systems the recently added pipe oprerator (
|>) now supports the use of the underscore (
_) as a placeholder on the rhs of the operator.
1:10 |> mean() |> round()
1:10 |> mean(x = _) |> round(x = _)
These two statements are equivalent, but
_ makes it possible to pipe by name the value returned by the lhs into any parameter on the rhs.
There are also some improvements to the ‘grid’ package adding flexibility to pattern and gradient fills and masks, which are recently added features. They are yet to be well supported by ‘ggplot2’.
Eighteen months from the previous release, this update brings many improvements to
autoplot() methods. These include enhanced capabilities for handling of normalized and scaled spectral data, as well as cosmetic tweaks to labels.
This update also includes smaller enhancements and fixes a bug. It tracks changes in packages ‘ggrepel’ (>= 0.9.1), ‘photobiology’ (>= 0.10.10) and ‘ggplot2’ (>= 3.3.3) and deprecation of functions in ‘tidyr’ (>= 1.0.0). The updated code depends on the revised
normalize() function in ‘photobiology’ (>= 0.10.10) and on ‘ggrepel’ (>= 0.9.1).
Version 0.4.4 fixes a bug affecting most of the geometries in the package. Adding multiple layers using the same geom to the same plot would result in only one of these layers being rendered with others silently missing from the graphical output.
Changes compared to version 0.4.3, the previous version in CRAN are:
- Fix bug caused by repeated grob and grob tree names.
Documentation web site at http://docs.r4photobiology.info/ggpp/ includes all help pages, with output from all examples, vignettes as well as a changelog in HTML format.
NOTE: Version 0.4.4 is on its way to CRAN.
Please raise issues concerning bugs or enhancements to this package through GitHub at https://github.com/aphalo/ggpp/issues. Pull requests are also welcome.