Sometimes I hate my analysis...

I'm actually kind of superstitious.  Every so often I really believe that if I talk about things I'll jinx them.  For example, I was heading to a basketball game with a few friends many years ago.  Golden State vs. Minnesota, so we needed to head over to East Bay.  About a third of the way across the bridge, one of the guys said he was amazed how little traffic was on the bridge at that time.  Within 30 seconds, we got stuck in traffic.  We were stuck on the bridge for nearly an hour.

I apply this sort of superstition to my thesis analysis.  I usually try to keep my deadlines for small (timescale of a day or a week) projects only.  None of the "my analysis should be done next month".  A little cynical, I know.  Lately I've been changing things around a bit, to the point that I was actually thinking I had everything done, and that I should have a completed analysis and be able to start trying to publish things in about a week.

Well, that was over a month ago now.

While I was finishing some things up, I noticed that there seemed to be one small little effect on my final results.  Tiny.  A couple percent or so.  Nothing to sneeze at.  I figured it should take me a day to run some sanity checks, maybe fix a minor bug or inconsistency in my analysis code, and everything would take care of itself.

Well...not so lucky.  It turns out this teeny little effect was evidence that I've got a bit of a bias in my analysis, and I've gotta get rid of it.  It's kind of a hard thing to nail down, and I wasn't really expecting it to be there.  It took a while to understand that it was a bias and not a bug.  Now I'm sure my code works, but I need to find a work around for this bias.  It might take me a little while...or I might get lucky and have everything work in a couple of days.

Anyway, I'm slogging through this stuff now, and it can be really depressing.  I can't move forward until I conquer this thing, so I'm kind of stuck until I figure a way out.  I hate being stuck...