What I'm up to this week (May 21, 2017)
Finally, got my computer back together just like I like it. Between that and trying to get some work done, I missed updating the blog last week.
Since the last update, I was actually able to meet with Darrell Pepper, and so I'm more motivated now to get the FEM conversion code to a cohesive, modern state. That was most of my work last week, although I'm getting a bit lost since I'm mostly approaching the code from a computer science perspective and not so much an engineering one. By that I mean that I can translate the code into a more rigorous, modern version, but I don't know if the code will still reflect the physical/engineering process it's meant to illustrate. And if the code is technically correct, but indecipherable to someone trying to learn finite element methods, then what good is it really.
To that end, I am now the proud owner of a copy of The Finite Element Method, Third Edition, the book all this code was originally written for. I'm hoping that by reading through it I'll get a greater appreciation for what the code is trying to model and so whatever I create will be of the greatest utility. And finite element methods are used for a wide variety of tasks, so a greater understanding of how they work wouldn't hurt either.
That will probably eat up most of my week, trying to at least get through the one-dimensional FEM material. But I do have Modern Fortran I'm still getting through, the unums project to work on, and a project with a friend of mine centered around baseball statistics (sabermetrics, if we want to be fancy). That last one has more work on R and SQL, two languages I've been meaning to learn more about, which is why they're getting prioritized. So, as always, my list for the week is:
- Get through one-dimensional FEM material
- Continue reading Modern Fortran
- Continue implementing Fortran unums
- Continue with sabermetrics edX class
If there's miraculously free time beyond this, I still have Probably Approximately Correct to finish, which is long overdue at this point. Once I get a few of these tasks knocked out, I do want to look into a short course on Julia for quantitive finance. If nothing else I'd like to get some more experience before JuliaCon, which is only a month away. Should be a good time!