Vacation

I'm in Ithaca visiting my parents and sister. While I will be studying this week, I don't expect that blog posts are going to be at the top of my to do list. Frankly, that's likely to persist throughout the summer. If I have something to say, I'll say it, but I'm not going to post daily just to indicate that I did a bunch of homework problems preparing for the Q.

Stats

The department finally published the content for the stats portion of the qualifying exam. Last year was the first year UMSL offered a concentration in Statistics, so this will be the first Q to cover it.

There are no big surprises: Basic Probability. Expected values and higher moments. Common distributions. Multivariate distributions. Random samples and the standard sample distributions (t, F). Sufficiency. Estimation. Hypothesis Testing.

Pretty much stuff that should be second nature to anyone pursuing a PhD in Stats.

Analysis

It's Wednesday, so it's Analysis day. Analysis has always been my best subject in math. While some find it tedious to prove that 1 <> 0, I find it comforting. Of course, I'm also aware that those same rules of logic produce things the Gabriel's Horn, a surface that can be filled with paint but not painted.

Just because math is logical doesn't mean it makes any sense.

Calculus

Calculus isn't technically on the Q. But, it's so foundational to all of mathematics that it seems reasonable to be brushing it up. The theory, I have no trouble with. Those little epsilons and deltas that give Freshman such fits do not bother me in the least. Instead, it's the stuff that the Freshman generally do fine with that I need to bring back.

This isn't a new problem. I recall very well that by the end of Freshman year, I was really struggling with Calculus. I could prove all the theorems, but couldn't get though the homework and exams because I hadn't bothered to memorize the formulas.

Memorization is one of those dirty words in modern education. The rote learning of a simple fact has been downgraded to a task that falls somewhere below intelligence. However, as a Chemistry Professor at Cornell once told me, "a big part of learning is remembering what someone tells you." I've since come to realize that memorization is much more than being able to parrot back a fact. It is that, of course, but it is also the ability to recall that fact in context. That is, it's the ability to realize that said fact is pertinent to the issue at hand. You can only do that when the fact is part of your working knowledge. If your brain has to fish around for it, you won't have it when you need it.

I recall a conversation with my father in the late spring of my Freshman year regarding Calculus. I was expressing my frustration that, despite understanding the concepts, I was getting hammered on the tests.

"Have you memorized the integral formulas?"
"Well, no, not memorized."
"Why not?"
"Ummm."
"Isn't that what the questions call for?"
"Yeah, I guess."
"You might want to give it a try."

I did, but it was so late in the term that I was only able to salvage a C in the course. Yes, really! A "C" in a Freshman math course. Not the stuff most math PhD candidates look back on fondly. That summer, I worked on a farm. That sucked. I decided I needed to figure this academic thing out. I was a 4.0 student for the rest of undergrad. Not because I suddenly got smart, but because I suddenly decided that memorizing stuff wasn't such a bad idea after all.

I still believe that understanding concepts is far better than memorizing facts. But, I also realize that there is some stuff you just have to know, period. So, it's Tuesday. I'm studying Calculus. And, in case you were wondering, ∫cosθ dθ = sin θ (arbitrary constant omitted).

Eigenvectors and values

Finally digging into what will hopefully become a decent study pattern this summer. Today's topic was Eigenvectors and Eigenvalues. As my text has a copyright from 36 years ago, the applications were predictably systems of Linear Equations, Quadratic Equations, and optimization problems. Of course, the text does touch on the result that any positive semi-definite matrix can be reduced to SDS^-1, but it's presented as a pretty theoretical result. Who knew in 1980 that one of the world's largest companies (Google) would be built on that result?

The Q

The qualifying exam will be given September 16. While I don't expect it to be as brutal as the one I took at Cornell, it's something to be taken rather seriously.

It makes for a potentially redeeming weekend. Right after finishing the exam, I'll be heading down to Berryman to run the Mark Twain 100 the next day. That's the only running race I've ever dropped out of. As with leaving Cornell, there were very good reasons to pull out, but it still feels like unfinished business.

So, in the space of 48 hours, I can assert that I really do intend to finish a PhD and assert that the DNF at Mark Twain was the fluke it appeared to be. Granted, I'm the only person in the world who looks at at either of those decisions to quit as a failure (well, maybe my Cornell adviser saw the first that way), but there is something very satisfying about going back and doing something right when the first try didn't work out too well.

Overtime

So, what's so important that I don't have time to write blog posts? Well, work, family, health for starters. Astute readers will note I didn't mention school; I haven't had time for that either. Wise readers will note that I listed those first three in the wrong order. I would agree, but I won't pretend that's been my reality lately.

This isn't unexpected, though it's coming a few weeks earlier than usual. Work usually shifts into overtime sometime in June and really heats up in July and August. The reason for this is that the primary initiative my group supports goes live every year on September 1. If that was the end of it, I'd be fine with it because everything in St. Louis heats up in July and August. Writing code in an air conditioned office isn't a bad way to pass the time.

Unfortunately, that's not it. September is usually pretty normal, but October and November, the two months when it really is nice outside, get busy again as the aforementioned initiative comes to a close and we get hit with a lot of support calls from folks frantically trying to get stuff done before the deadline.

I've tried to inject a little sanity into the process. I'm finding this a bit easier in my new group. We're smaller and our part of the system, while it has it's challenges, generally gets fewer changes from one year to the next. We're putting out some big efforts right now, but there's every reason to believe that we will be finished with development in the next few weeks.

That's a good thing because, for reasons that are unknown to me, the powers that be are planning on turning this thing on in early August rather than waiting for September. That's compressed our schedule a bit, but not tragically so.

This is what I signed up for. I had the good fortune (though that's not how I viewed it at the time) of working on a very high-stakes, deadline-driven project before I even got through undergrad. I realized that this is just how the industry works. I could either embrace it, resent it, or find some other line of work.

I chose the latter option, leaving engineering to become a bike racer. I learned that even "dream" jobs like being a pro athlete have some pretty serious negatives. But, I also learned that if you can get past that, it's possible to derive tremendous satisfaction from almost any job as long as it's something you have a passion for and are willing to put out some effort.

And, it's been mighty useful for this whole grad school thing to be able to look at the effort required and realize that I know how to put that out. So, I'm not complaining. I just have a lot to do.