Tammy Jackson, a senior research statistician developer in the economics technology group, says her job is like taking a math test all day—but she wouldn’t have it any other way. With a background in matrices, linear algebra and math education, Tammy is proud of her work and has a keen eye for program flaws. Read her interview to find out more!
Can you talk a little about your position at SAS?
Well, they used to call us PROC writers, meaning we write SAS procedures. I work specifically with econometric time series. When I deal with time series statistically, I deal with what are called discrete time series; this means that the time is measured in monthly, weekly, daily or hourly increments. The problem is that this is divorced from continuous time, and going from continuous, real-world time to discrete time for statistical purposes isn't easy.
What is your educational background?
I went to NC State, and when I was there, I was borderline math and computer science. But, I decided that I didn’t want to sit behind a computer all day (And, yet here I am!), so I picked math. I double majored in math and math education, and I was going to teach math. But after a while, I decided that I couldn’t deal with the bureaucracy of the schools, even though I loved teaching. I had always wanted to work with scientific applications, so I did computer graphics programming for GE aircraft engines.
When we moved back to Laurinburg, NC, I started going to UNC Pembroke for their Masters in Math and Computer Science Education program. I may be the only person to have ever graduated with that degree! At that point, I’d had so much programming experience that I pretty much knew everything I was being taught. After that, we moved to Roanoke, VA, and I ended up getting a Masters in Applied Math from Virginia Tech. When we had the opportunity to come back to Raleigh, which we really wanted to do, I started teaching CS programming at Wake Tech, and I began working on my PhD at NC State.
How do you use math in your job?
During my studies, my areas of expertise were numerical analysis, linear algebra, computational math, matrix theory, those kinds of things. We solve linear systems and non-linear systems, but even when we solve non-linear systems, all we do is take a linear approximation and solve that. Just about everything I do comes down to using an iterative, linear solution, meaning I set up a big matrix to solve the problem. When I end up with a problem, I’ll sit there looking at numbers fly by for three, four, five days straight. These are problems that require thousands of matrix calculations.
Essentially, my work is like sitting there and taking a math test all day—a lot of people can’t do it or don’t want to do it, but I do.
Do you think the environment at SAS is conducive to mathematics?
I think it’s something that sets us apart quite often from other companies. Right now, I’m looking at a problem that’s off by half a percent. From experience, I know that’s just big enough to tell me that something’s wrong, even though the casual observer may not realize that. If the difference was more like 10-6, then I’d say it’s just a floating point difference, but I know that with half a percent off, there’s some small flaw in the program. And even though I could probably get away with keeping that flaw in and perhaps never get caught, I just couldn’t do that. It wouldn’t bother a lot of people, but it bothers mathematicians. It speaks to SAS’s integrity as a company.
Have you done any cool work in mathematics?
I’ve got one patent and a couple more in the works, so I’m on the patent wall. It was for PROC HPFEVENTS—it’s a method of parameterizing regressive features in time series. Essentially, it’s a way of describing things like promotional sales or holidays. Before this, everyone had to make their own series, which was very tedious. But this automatically generates a series for you.
What about math appeals to you?
My dad was a high school math teacher, and my second grade teacher actually lived in the same house as me, since it was a teacher’s apartment. I could just go down to the end of the hall and knock on my second grade teacher’s door! I remember her teaching me how to multiply and divide in second grade, and I just loved it. I was five years old and I loved it, and I’ve loved it ever since. For me, it’s just always been there.
Who is your favorite mathematician?
I like Pierre de Fermat. Fermat’s Last Theorem was solved several years ago. The story behind that is that he wrote in the margin, “interesting proof, won’t fit in the margin.” One of my professors said that if you’re really a mathematician, once you know the answer, you lose interest in the problem. And that’s true! It’s hard to explain, but once you know the answer, it’s like it isn’t even worth writing down. What intrigues you is getting there.
Do you have any other hobbies you’d like to share?
I exercise a lot. I jog five miles a day. I beat my son and son-in-law a couple of weekends ago!
What kind of advice would you give to students studying math?
Learn to solve problems. I’ve taught math and that’s always the biggest mistake. Students seem to think that if they look at a problem and the answer doesn’t pop into their head, then something is wrong. No, no, no! When I teach, I try to show students where I start, what’s going through my head and how I’m solving the problem. That’s one thing. The other is: don’t stop taking math until you’re absolutely certain you’re at the level you’re going to need. It’s very difficult to come back to math after an interruption.
If you’d like to chat with Tammy about her interview, feel free to contact her at firstname.lastname@example.org.
If you know someone at SAS who really loves math, nominate them for an interview! Just email me at email@example.com.