So all week, I have been working on trying my best to understand Bayes Theorem and Belief (Bayesian) networks. My background in statistics is pretty pathetic, so I’m playing catch up before the semester ends. I’ve seen Bayes before, but only as a general classifier; I never had to actually work out the probabilities and CPTs to the belief networks.

Anyways, since I haven’t posted in a while, I have been wanting to drop a heartbeat. Thump thump. I also wanted to mention an article that has saved me so far dozens of man-hours finding bugs. If you deal with native applications, you’ll find gdb to be an indispensable tool. I found this article on developerWorks that showed me a few tips and tricks to harnessing the power of gdb.

Until I found about conditionals and watches, I had been using gdb like a caveman! Rather than wading through a function call being called correctly a few hundred times to find the one broken usage, conditionals can be used with your breakpoints to filter out all of the normal situations. I also found watches to be useful to see when and where a variable / memory address is accessed or modified to see what code might be breaking things.

I had to work through something where a function call works 99% of the time and many things have to be done to the application / system before the problem manifests itself. Rather than setting a blanket breakpoint on the function call, I sifted through for when a parameter matched a certain unwanted value or the function reached an undesirable state by applying a conditional on the end of the watch or breakpoint. Thankfully, IBM wrote a clear and concise article on the basic powers of gdb.

This week is the last week before the final. I have a 1 week break before the qualifier, so things won’t be so rushed this semester. Since I have some time to kick back and catch up on my rest, I think I will take up MensJournal on their article about the top 25 beers and enjoy as many of these as I can this week.

In fact, I’ve added beer as a regular part of my diet. The good news 3 beers a week (post-run drink) help add a little bounce in my stride. Unfortunately, I have put a few pounds back on. I really like beer and would hate to completely give it up. I might have to limit my intake to only after long runs on Saturdays.