So you want to be an astrophysicist? Part I - undergrad

Part 2 of Ye Olde Blog So You Want To Be an Astrophysicist? series.
Lightly re-edited.

Should you do astronomy as an undergrad? (the following is in part shamelessly cribbed from prof Charlton's freshman seminar for our majors):

• Do you like stars and stuff? If not, you probably should look for an alternative, on the general principle that at this stage of life you should at least try to do things you actually like.
  If you do, good for you. Now, do you have the aptitude?
• Professional astrophysics/astronomy is not about looking at stars (except at occasional star parties, for outreach or as a sideline hobby - and a fair fraction but by no means all astronomers are enthusiastic amateur astronomers). Nor will you need to learn about constellations, or speculate about the meaning of it all, or the origin of the universe or other sophomoric philosophical issues (except over occasional beer sessions - except for the constellations bit).
  You don't need to be an astronomy major to speculate about the meaning of it all over late night beer.
  What you will need to do, is at least 75-80% of a physics major (and preferably all of it, physics double majors are a common path, as is just doing an all physics or math/physics (like me!!!) path, and adding astro later). That's four years of 2 classes per semester, calculus based physics. You will also need at least 3 years of university level calculus, and if you find yourself taking as little math as possible, then your career options will rapidly shut down and you might want to rethink. Some computer science or electronic engineering wouldn't hurt, though most of the practical computing you need you will be expected to pick up through self-study. So, you would need, for example, to be able to look at HTML sample code, or a "how to web page", or in a pinch a book, and figure out in few hours or days how to do adequate HTML coding, as a minimum. Most astrophysics types are expected to know one major compiled language (C++ or Fortran most common), several macro/mark-up languages (like TeX/LaTeX, IDL or Perl) and higher level languages as needed.
  Most people find this to be hard work. You should be ready for hard work.
• Do you like to read? Cause you'll be doing a lot of it. Books, papers, web pages, class notes; and, whether they admit it or not - science fiction (ok, not all astro types are science fiction fans, just most of them, secretly, open Trekkies, whatever. They are of course All Hot. And know better than to refer to skiffy as "sci-fi" [sic]).
  What do I mean by lots? (For an undergrad.) Mean output of a professional astronomer is 3-4 papers per year. Each paper has 30-40 cites to the literature on average. You have to have read those, all of them! Now, if you work in a single sub-field (not uncommon) there'll be a lot of overlap between cites in successive papers, but you'll also have to read 2-3 papers for each one you cite. And, you need to keep up with the literature, new papers every day... So, we're talking 1-200 papers per year.
  For starters, as absolute minimum for continued professional existence.
  For ever.
  Enjoy.
• Exams. Yep, we have those. Some people can't handle them. Don't know what to do about that, they include brilliant people. Same with essays and projects. Different people can't do those. Don't know what to do about that either.
• Research. It is generally a good idea to try to get into some research if you can, typically summer after the end of your junior year, earlier if you can. It looks good on a resume, helps you get letters for grad school (if you still want to go) and lets you know if research is the sort of thing you want to do. Some people hate it. Better to find out before you spend ~ 5 years in grad school. That's a big opportunity cost if you just want to go out and earn money (or a wonderful life experience if you have the luxury to have those). Some people hate doing research.
• What university should I go to? Well, the one you can. The "best" one if you have a choice.
  Does it help to go to a "name" university (top private, Big high profile State, or a high rep liberal arts college) - you betcha (analogous situation for other countries, I know the deal for some, not others).
  What does it buy you? 1) a shot at a good education. In the US there are maybe 50-100 universities where you can get a very good education in astrophysics; the other few hundred are "good in parts", but your odds go down sharply. 2) a second look at the next stage, the committees will look twice at people from places they know (and conversely if you do badly at a known university, they know too. But, if you do badly at a university they don't know, then you will not get a look at grad school). 3) at some level you get what you pay for - there is not a perfect correspondance, but high correlation.

  So - the top private places (Ivy Leagues, Stanford, Caltech, MIT) - will give you a strucured, superb education, with access to top faculty (half of whom may not care about you). For a price. If you're willing to take advantage of it. And there's significant in-house competition. Not matter how good you are, you're going to meet someone better at those places.

  Big State: well, they're relatively affordable, they are big, so there's a broad range of courses and people, and they're well enough known that if you do well you'll get a look for grad school. The catch - you have to self-motivate, the place is big and you can all too easily vanish. Embed yourself in the department, not the dorm or the frat/sorority. Interact with the faculty, go to talks, talk to people. Profs are PAID to be there at office hours, and without exception they have a unique weakness, they love talking about "their research".
  Suckers.
  So go bug a prof. Lots, repeatedly. Oh, and get good grades and have natural math aptitude. If you do well academically (B+ GPA or higher AND a consistent good performance in the "hard" classes) and if you hook up with people to work with, you're set. But YOU have to take advantage of the opportunities. If you're lucky someone will reach out to you once, maybe twice, but beyond that it is up to you.

  Lib Art: small, fabulous teachers (on average, at the good ones, except when they're fabulously bad) and lots of personal attention. A disproportionate fraction of top researchers come out of majors at these places (like Reed, Vassar, Swarthmore etc) and they produce a disproportionate fraction of science majors, compared to their size (US state universities have an appallingly small fraction of science majors, and vanishingly few physical science majors). Catch, there a a lot of mediocre and bad ones...

  A bad university, in the physical sciences, is a career end, unless you have exceptional persistence. What to look for is a sequence of hard calculus based classes - look for advanced electromagnetism, or advanced quantum, with 2 physics prereqs and a high number math dept calculus based prereq. A lot of places let this slip, and you get soft, non-calculus based classes filling the major requirement. Fun, but you don't learn the basics you'll need later. The US "modular" course structure is particularly bad in this respect (more on that in another thread, but think about how much more in-sequence engineering and other serious professional majors tend to be... I'm biased, I was educated at a UK university. Viva la difference).

• Publishing a refereed research paper as an undergrad always helps, but is not essential. Getting into a national summer intern/REU program helps, but is not essential. Having good solid grades, and doing well on the GREs is what matters.