If you're insistent about sticking with scipy's gnuplot, I've found that by far the best way to use the interface, and to unlock every gnuplot option, is to use the "gplt.load" command. You first need to create a file with data you're going to plot, formated exactly as you would format it to bring it into gnuplot and plot it manually. Then write a file that contains the gnuplot commands that you want to run, as you'd type them on the gnuplot command line. Finally, from within scipy execute the command gplt.load('filename.txt'). This will load gnuplot and run the commands as if you had entered them in the gnuplot interface. Its not the most elegant way to interface with gnuplot, but between the simplistic interface of scipy's gplt and this work around, you can create any kind of gnuplot you want without having to learn another package. Generally I'd recommend using the simplistic gplt interface to draft your graphs, then when you know how you want them to look, or need to add features gplt doesn't support (like error bars) you use the method I described above. I've written some routines to automate some of the above described methods, though there is still one more modification that I'd like to make. However, if you're interested, I'll upload it to some web space so that it can be downloaded. -- R. Padraic Springuel Teaching Assistant Department of Physics and Astronomy University of Maine Bennet 214 Office Hours: Monday 10:00 - 11:00 am; Thursday 12:00 - 1:00 pm