Here's a handy command:

whereis FILE

Great for searching recursively for a certain file; a bit like "grep" for filenames instead of file contents.

I figured out how to change the fonts on the graph. I had to make sure that I specified at the beginning of the file that I was going to use GD and GD::Text, and I specified where the program should look for installed fonts. Since I'm using strict, if I don't say "use GD," I get a (misleading) error message about using barewords with strict and so on. Anyhow, the graph can now use system TTFs or built-in GD fonts, and the graph is a bit more readable now. I do need to have the graph adjust when there is a lot of data, because the labels overlap and are unreadable.

I removed the "order_by" bit from my query, since it actually wasn't affecting the results and could confuse someone (for example, me!).

The graph shows how many times the sensors logged a certain level of restlessness while a person slept. Problem is, every time a restlessness level of 4 is logged, levels of 1, 2, and 3 are also logged immediately before. In other words, it has to go through all the lower levels first. This can distort the data towards the lower levels. I wasn't sure I was completely clear on the concept, so I checked through the data, and it almost never violates this rule (less than 5% of the time in all cases I searched for, and actually less than 1% in most cases). It's possible that most exceptions can be attributed to a missed data signal, since I'm told that the X10 wireless protocol doesn't do handshaking*. I thought about trying to correct the data by tweaking my database query, then decided that that was too complicated and just processed the data programmatically after it was pulled. The graphs are now correct.

I showed Chinonye the virtual agents I found and sent her their sites.

*coordinating data transmission so that it can confirm that the data it sent was successfully received