Getting Matplotlib to meet the maniacal figure standards of my grad advisor.
My grad school advisor is a great man and a brilliant engineer. We should all be so lucky as to find the passion for our work that he has for his. It's a privalege to work for him.
But the way he makes plots totally sucks. I won't name the software, but let's just say that it makes me wish I could use excel. He uses it because he likes to tinker with plots: add arrows on the axis labels, put annotations all over the place, etc. It's excessive, IMO: the annotations detract from the often-significant message of the plot. But that's his taste, so it is what it is and I don't fight it. But Kaleidagraph...oh man. I'll fight it til the day I die (or graduate, whichever comes first).
Motivated by my disdain for this POS software for which the most recent version was release in the last 5 years but still looks and operates like the original version from 1990; motivated by the hope that if I could make identical-looking figures through some other platform then I could stop using that pathetic program; and hoping to just learn a whole lot more about matplotlib... Here are some style sheets and functions which get us really stinkin' close to his standards.