[Rd] changes in upcoming lattice

From: Deepayan Sarkar <deepayan.sarkar_at_gmail.com>
Date: Thu 14 Sep 2006 - 18:18:31 GMT


Hi,

there will be some fairly major changes in the lattice that will get released with R 2.4.0. A first version is now available on CRAN, at

http://cran.r-project.org/src/contrib/2.4.0/Recommended/lattice_0.14-3.tar.gz

Although there is a dependency on R 2.4, the package passes R CMD check on 2.3.1 as well if you change the Depends line in DESCRIPTION. In other words, you should be able to test it even if you are not running r-devel.

There are 3 types of changes:

  1. Some new features. These are of the most interest in the long run (albeit to a very small number of users) , but they have not been thoroughly tested or documented (and some features not even implemented). Documentation, such as it is, can be found in

?axis.default
?packet.panel.default
?panel.number

2. Bugs that have been introduced in the course of implementing these new features. Of course I'm hoping that there aren't any such, but I'm sure there are. So testing your lattice code to flesh out these bugs before the 2.4.0 release would be appreciated.

3. This is where I might get some legitimate complaints. For the last several versions of lattice, there has been a feature where the panel function would get passed arguments called panel.number and packet.number under certain circumstances. This was to enable the panel function to know which panel was being plotted. This was intended as a quick hack, but apparently several users have found the feature useful. Some new functions, described in ?panel.number, address this issue in a cleaner and more general way.

Given these new functions, I would like to remove the earlier features (arguments to panel functions). In fact, I have removed them in lattice_0.14-3. I will put them back if other packages on CRAN start to fail or if enough users want it back. I would normally have kept it in with some sort of warning, but the formal deprecation mechanism doesn't work in this setting, and I would rather start cleaning up code that made this work than make it even more complicated. I would be happy to assist package developers who want to write portable code that works in both older and newer versions of lattice.

Comments / suggestions / bug reports are welcome.

Deepayan



R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel Received on Fri Sep 15 04:21:17 2006

Archive maintained by Robert King, hosted by the discipline of statistics at the University of Newcastle, Australia.
Archive generated by hypermail 2.1.8, at Thu 14 Sep 2006 - 19:30:08 GMT.

Mailing list information is available at https://stat.ethz.ch/mailman/listinfo/r-devel. Please read the posting guide before posting to the list.