Currently, most work on theobio is done using a handful of programs:
- R is a general-purpose computation and statistics language
- gcc is a compiler for the low-level programming language C.
- perl is a flexible scripting language
- python is a generally agreed to be better than perl, but we haven't mostly switched yet
- Octave is a free, matlab-like program.
- We are also thinking of adding some bioinformatics functionality.
- We have started mucking around with PAML
- Gnuplot is a flexible plotting program
- We are developing rules for gnuplot on a project here
Stuff in the default makefiles
Most of this is not documented, or documented but not easy to find. I'm trying to catch up.
Our R machinery is very general, and a little bit complicated. I have the idea of making a parallel track ((R step)) where each script is designed to only make one target (although it can still contribute to multiple targets, via having its final environment read in).
ls lists the working directory.
%.ls passes arguments to ls.
The <code.compare target can be a useful failsafe when debugging.
There are some customizations we make to skins, extensions and other components of our MediaWiki installs.
- Some wikis use a custom skin called Stealth Ninja (working title), which hides wiki interface elements and presents a clean interface.
- Extension management
We have a hacky but working Refs project that can be used to make links and bibtex files (to be used with latex) from Pubmed IDs or links.
Many projects on theobio use "dumb markup" ( dmu), basically random assortments of rules for source files that are supposed to be human-friendly. We eventually hope to regularize and document this.
The main settings file is /usr/local/mediawiki/wwfarm-1.21/template/extensions/enable-ww.php