On 7/21/06, David Wooten wrote:
Greetings all,
I've been itching to stop using "Illustrator" and other programs to make the charts and graphs I wish to include in my ConTeXt-generated documents, and would like to get some advice on the matter.
There is a special (albeit not well furnished) place in my heart for Tufte's /"Visual Display of Quantitative Information/", so I suppose I am always leaning towards that philosophy of chart-making. I came across Jean-luc Doumont's article /"Drawing effective (and beautiful) graphs with TeX*" [1]/, in which he refers to his macro package called JLdraw---and shows some pretty examples. JLdraw was in fact never generally released, although when I contacted him he was happy to send along the macro package. In beginning to tinker with them, I have not had much luck getting them to work within ConTeXt, undoubtedly due to my persistent naiveté.
Thus I'm curious as to what others use... is R an efficient method to produce elegant charts? Is straight MetaPost preferable?
With metapost you can surely achive most beautiful results and it is not as difficult to learn as TeX-programming. Of course you might need more time to draw what you need or to write your own set of macros, but if you have high demands about quality this might be the way to go. However, if you prefer doing it quicly using the existing tools (be aware that you have to learn how to use those tools as well), R or gnuplot might be an interesting choice. You'll be limited by the power of those two tools, but in most cases they should suffice for the normal usage. The gnuplot module is still in development (I've been just begging Hans for help a few hours ago ;). Take a look at the demo section of gnuplot (http://gnuplot.sourceforge.net/demo_4.1/) to see if it can offer you what you want to do. In that case ask on the list again, I'll give you further pointers how to use it with ConTeXt (http://pub.mojca.org/gnuplot). But basically you can take any program to draw graphs and include the resulting PDFs. I'm afraid that the macros from the paper which you pointed to, use some PostScript code that cannot be handled as-is (you need some conversion to PDF first) and I'm affraid that the effort put into making it work woudn't pay off now that you have a great varienty of other plotting programs, including metapost itself (esp. if the package has never been released - you'll probably hardly get any support for it). Mojca