Wednesday, April 13, 2005

My own little hell.

I just got back from the library where I checked out a book entitled Discrete Multivariate Analysis: Theory and Practice -- I'm hoping it's more practice and less theory. When I asked the librarian whether I was allowed to check it out, she replied, "Yes, but why would you want to?"

I do not know what "discrete multivariate analysis" is, and I have had very little training in mathematics (some calculus, some statistics, and some other stuff I picked up along the way). There is no chance in hell that I'm going to understand all of the material in this book. I'm hoping the little bit I need is somewhat clear to me after reading it over twenty to thirty times.

If the material is totally out of my ballpark, why would I ever want to check out this book? It all started with these papers published over fifty years ago. I began reanalyzing their results, and that led me to this paper on fitness and recombination. I am now trying to apply a statistical test developed in that paper to a more complex scenario. To do so, I need to figure out how a value in that paper was derived so that I can derive it for my example. That's where the book with the scary name comes in.

Last time I checked, I'm working on a degree in Genetics. I never thought I'd be getting books from the library on mathematical concepts I didn't even know existed when I signed up for this gig. I guess no scientific discipline exists in isolation, and in order to study biology in this modern world you need to have a good handle on the life sciences as well as math, physics, chemistry, and anything else that may intersect your path.


At 4:32 PM, Anonymous Reed A. Cartwright said...

You're not just studying genetics, you're studying population genetics. It is because of population genetics that modern statistics and multivariate analysis exists. Fisher had to invent modern statistics to deal with the compex data that he had to deal with. The physicists are correct; biology is not like physics. We wish we had such a simple discipline.

As a population geneticist who specializes in math/comp bio, I'll be happy to help you out with your stuff.

At 4:56 PM, Blogger RPM said...

Point taken, Reed. It's always fun to juxtapose the the theoretical quantitative aspects of genetics with the more molecular research. I've actually considered exploring some molecular biology questions as well, just so I have a thesis that's even more all over the place.

Of course, calculus was only developed by Newton so that he could understand the decipline that came to be known as physics.

At 6:06 PM, Blogger Razib said...

i am reading fisher's biography (written by his daughter) right now. in any case, i remember there was a story about some mathematicians who read about a talk that the 'famous biologist r.a. fisher' was giving in the paper, and they looked at each other and were like, 'is this the same r.a. fisher who is a famous statistician?'

At 11:24 AM, Blogger Ranger of the West said...

“I occasionally meet geneticists who ask me whether it is true that the great geneticist R. A. Fisher was also an important statistician.”

Leonard J. Savage Annals of Statistics (1976)

A Guide to R.A. Fisher

At 2:29 PM, Blogger RPM said...

Just to clarify (in case my rant was taken the wrong way): I did know about the quantitative aspects of evolutionary genetics when I started my PhD, and I do actually enjoy some of the mathematics involved in population genetics. Hey, I've even TAed a population genetics course! I was just kinda suprised to be delving into (what seemed to be) such abstract concepts after beginning with an entirely biological question. And now, after looking over the book, I can extract what I need without too much blood, sweat, and/or tears.

At 12:04 PM, Blogger Sya said...

"I guess no scientific discipline exists in isolation"

I've actually thought this was a really cool thing. Who needs to actually pick what field to go into when you're going to apply a lot of different things to whatever you're going to study anyway?

At 3:44 PM, Blogger RPM said...

So, the reason I checked out the book was to figure out how to extend a three dimensional analysis into a fourth dimension. If you thought drawing a 2x2x2 table on a sheet of paper posed problems, imagine how much fun it is to draw a 2x2x2x2 table. I'm now at an impasse, as it appears that I have no idea how to calculate the cross product of the 4D table. At least, I think I'm calculating cross products -- as you can see, I'm about thisclose to illiterate when it comes to math. What I'm doing is multiplying accross diaganols (are these cross-products?). I've tried two different times to come up with the answer, but neither one turned out correct upon double checking my work. I'm going to stop bitching right . . . about . . . now.


Post a Comment

<< Home