Thursday, January 18, 2007

Mauling the language

My research involves doing a lot of computer simulations; I write nearly all of my programs in (get ready to be disgusted and swear you never read this blog) Fortran. This morning, I've been spending time trying to adjust some code so that I can compile a program on a few different machines - they're each persnickety in slightly different ways, and I have to be careful with different issues. If I can get it to work, it means I can use more machines, though, and therefore run my simulations faster. This, of course, will free me up to do even more simulations, which means more work, which means... oops, better not think about that right now.

In any case, I've been banging my head against one particular platform for the last hour or so, not sure why my program was giving funky results. I just realized, though, that I had a typo in a crucial function. While trying to multiply two matrices, instead of using the "matmul" function for matrix multiplication, I was eagerly encouraging the computer to "matmaul" the data. This may have been my first digital Freudian slip.

No comments: