Progress Report for the week of November 11, 1997

Accomplishments:

This week was one of the weeks where very little when right with much ease. We spent a lot of time doing research into various Java API's. Matt is well along in getting parallel coordinates to graph correctly. It graphs, but the graphs don't look right for some of the test data sets, so we have some debugging to do. It could be in the graphing algorithm or in the method that reads data in from the file. Matt also spent time replacing java AWT components with the equivalent SWING components. We've spent time trying to figure out how to compile with the SWING components on crane. I've had no luck. The swing library comes as a .jar (java archive file, similar to a tar file). The only way I've gotten the program to compile on a non-windows platform is to unjar the file. In windows, the .jar file just has to be in the CLASSPATH. That doesn't work in the unix versions I've tried (LINUX, Digital UNIX, and Solaris). Unfortunately for us, we can't have the file unjarred on crane since we don't have enough inodes allocated to us. There are about 800 files in the archive and we only can have 400. For now, we are going to do a lot of the development on our own machines. Working on this problem took up much of the week.
We got together for our second code review. I went over how I did the graphing of the parallel coordinates graph. Once I figured out how to display the graph object on the screen, I just had to perfect my algorithm of how to plot the lines. In the code review, I went over the techniques I used to access the data points so that Will and Jason can develop the plotting algorithms for scatterplots and glyphs. Once they perfect those algorithms, then we will have two more working graphs.

Questions:

To do List:
BACK to MQP homepage