Decisions, decisions...

[geekusprimus]

I'm planning on applying to graduate school this coming fall, and I'm trying to get a better idea of what I'm interested in to help me as I look for schools. I've done a lot of undergraduate research in relativistic astrophysics from a computational side, and I find certain aspects of it interesting enough that I think I might enjoy continuing along the gravitational physics road for graduate school.

Now, I see two real routes I can go:

• I can stick with doing computational stuff and focus on numerical relativity, which has become a really big field over the last few years (in no small part due to LIGO). I already have a decent amount of experience with numerical methods and a lot of the questions in this field, but I'm really concerned about becoming a code monkey. Apparently I have a knack for debugging numerical codes, so that's what I end up doing most of the time when I would rather be tackling the physics questions. Some of it's pretty interesting, like making pretty pictures or searching for physical constraints and restrictions on a system so we can make better estimates and improve the accuracy of our simulations, but a lot of the time I'm just cleaning up someone else's mess (like optimizing/rewriting badly written packages so our simulations can run within the expected lifetime of the Sun).
• I can transition into a more analytical route that leads to things like theories of quantum gravity, black hole thermodynamics, or other topics like that. My concern here, though, is that I'll end up in a group somewhere that is more fixated on the mathematical intricacies than the physical questions. I attended a gravity conference a couple months ago and noticed that an unfortunate number of my peers at other universities were presenting not on physics problems, but rather math problems, like the person who spent nearly an entire 18 minute presentation just discussing whether or not a particular equation was meromorphic, or the student whose entire research was basically exercises in group theory applied to the Einstein equations (which were too mathematically trivial to interest the mathematicians and too lacking in physical connections to excite the physicists). I'm sure these are important questions to someone, but there was absolutely zero attempt at explaining the relevance of these mathematical results to physical problems, and that bothered me a little.

(Experimental/observational gravitational physics has also exploded somewhat since LIGO, but I'm a pretty terrible experimentalist, to be honest, so that's not really on my radar.)

Long story short, I'm worried that I'll either turn into a code monkey or a bad mathematician. For those of you who have experience (or know someone who does) in computational or theoretical physics (not necessarily just gravity), what has it been like for you, and how likely is either outcome? What have you enjoyed about your field, and what has been difficult or not particularly pleasant about it? Any additional perspectives or insights would be welcome.

[Nishikata]

I did have this kinda feeling last year, when I can always find something that I don't like in every possible field. Now I am more accepting of the interesting/boring parts of the work, and try to keep entertained on the interesting part as much as possible.

Some of it's pretty interesting, like making pretty pictures or searching for physical constraints and restrictions on a system so we can make better estimates and improve the accuracy of our simulations, but a lot of the time I'm just cleaning up someone else's mess (like optimizing/rewriting badly written packages so our simulations can run within the expected lifetime of the Sun).

In all collaborative work, [someone's mess] always exists. Sometimes we clean theirs, sometimes they clean ours. So I suppose you will be alright if you have them balanced, and thus choosing the PI that is responsible and engaged with the team is a good step to have the mess in control.

I'm sure these are important questions to someone, but there was absolutely zero attempt at explaining the relevance of these mathematical results to physical problems, and that bothered me a little

Some people are really not good at presentations, but they should always be good at grant-writing or they won't be able to continue their research.
When people write grant applications, they write in a very structured manner that all steps will lead to something, and that something is significant.
My suggestion is when you encounter that kind of presenters, talk to them privately after the session to ask how is meromorphic equation relevant to the physical system, how they think is important and so on. They might be working on minute improvements, but never a zero-relevance topic to the field. It is perhaps just so obscure that we can't get it yet.

In choosing between the two, I'm more inclined to do computational stuff. Maybe it's because I am not a strong mathematician, or maybe it's my survival instincts telling me that computational stuff would allow me to transit to other fields more easily if things don't work out here.