How do I transition from physics and math to programming?

Question

I'm a physics PhD with little actual programming experience.

I've always liked programming and played around with BASIC, Pascal as a teen, but the extent of my experience writing complex programs comes from an introductory course in computer science.

Now I've decided that I'm more interested in programming than in physics and started to learn Java.

Coming from a physics or math-heavy background, what would be the best strategy to maximize my value in the field?

Answer 1

Have you considered being a Physics person who know how to program? Math heavy programmers are rare.

If you are still in university, take the introductory programming class. You will get a good idea of how more rigid programming is to be done, and allow you to make the decision on a more qualified basis.

Answer 2

IMHO, your field of study (Theoretical particle physics) + Java or any other programming knowledge gives you a huge advantage. How about learning (basic or even advanced) Linux, distributed computing, algorithms, etc? You certainly need Unix/Linux skills for most scientific research projects. Many of my friends from other disciplines (Polymer, Mechanical, Industrial, Electrical engineers) are also good computer programmers (C, C++, Java, Python, etc) . Few guys are now studying in Uppsala, Sweden :)

How about looking for all science related computer jobs in universities, scientific/research labs, government/private science labs, electronic product companies, pharmaceuticals, defense, etc? AFAIK, EU has a lot of particle physics labs. Are you interested in expanding your search to those countries/labs? I guess you might have done that already. May be, one more try?! Physics or Science or research oriented IT jobs are your best bet.

If you want to jump into the IT programmers pool, you have many options, based on the job market in Sweden. I don't know the details of Sweden IT market. Big corporate jobs look for Java/J2EE experience; please look at the skillset they want. You need strong Java & J2EE skills, Spring, Hibernate (ORM), Application servers (JBoss, Websphere, Tomcat) and other related topics. For mobile companies, you could learn Android. Your knowledge of particle physics may be helpful in few conversations, I would hope.

You could also try consulting jobs by joining one of the IT consulting/recruiting companies instead of full time employment. Send them your resume and see what they can offer you.

Good luck!

Answer 3

I'm a Java programmer with a math Ph.D. I started programming at age 36, so our situations are similar.

I think that you will find that your degree will open up doors for you even in non-physics, non-math situations.

Since where I work is important to me, I haven't tried to maximize the math angle, though I may try to get into some financial/banking programming eventually. I've stayed away from science research, as I expect that sort of thing is not so broadly available. (As in, plentiful jobs in every city.)

My recommendation is that you learn the basics of web programming, get your first job on the basis of your certification and being a smart guy, and then work on learning to be a really good software developer for a year or two. At that point, revisit the question of who wants a programmer that is really good at learning physics/math.

Here's how it has gone for me so far: My first job was at a startup as a full-time intern, making Java web apps. After nine months, I got a job as a consulting with a very good IT consulting firm. I've been with that company for nine months, and it is going well.

I haven't yet taken the next step of looking for a position that incorporates my math background, for two reasons. I don't want my resume to seem to indicate an inability to commit to a company for any length of time, having three different jobs within two years. The other reason is that I'm finding my compensation increase fairly quickly in the IT consulting world, so I'm not in a particular hurry to leave.

Answer 4

Your core strength compared to other developers is your math and physics background. I would try to stick to fields of software development and technologies where that is an advantage, rather than for more general programming. For example, you're probably going to be more successful in algorithm development, robotics, or defense, or finance, than you are going to be in general web UI programming (where folks with a design background might have an advantage).

Have you considered working in the financial software or algorithmic trading fields? I've known my share of hard-science folks who made this transition fairly well, better in fact than pure CS folks because of the stronger mathematical background. Many of these jobs seek PhDs rather than extensive programming experience.

Answer 5

I'm a Java programmer. You can pretty much teach yourself programming, Java or not, and most great programmers are self-taught, I'd say. This means that the entry barriers to programming is very low, which leads to over-supply of programmers. Results? Lower wages.

I'm not sure if you want to make a career change at this point in your life, especially when you're an expert in your field. Judging from your credentials, you're not exactly young. You need at least a couple of years to learn programming, not just Java syntax or how to use libraries. So by the time you're ready to work as a programmer, you'll be a couple of years older.

There are tons of kids, fresh out of college each year who are mostly younger and better at programming. With everything being equal, employers in general prefer younger people because not only do they learn quicker, but more importantly they are easier to control and will settle for less salary.

Physics phd's are a lot harder to come by than Java programmers because the entry barriers to phd degrees are a lot higher. With your planned career change, you're giving up your comparative advantage as a physicist. Instead of dropping your physicist career to become a programmer, supplement it with your new found interest in programming. You'd get the best of both worlds that way. Like other posters said, math or physics savvy programmers are very difficult to find. You'd be invaluable as a phycist who can communicate with programmers.

Answer 6

Thats what i am ... well i learnt c# instead of java (and id encourage you to do the same as its a far larger framework that is still being evolved at a rapid pace - although i dont know what is prevelant where you are.).

You need to make a decision though ... do you want to be a programmer or a physicist that programs?

If you want to be a physicist that programs you probably want to look at C or soemthing functional. Wheras if you want to be a programmer that can quickly write apps then java/c# is the way to go imo.

Very different subjects - i have abandonned physics and kinda miss it tbh as i cant get work in it. Now i am a developer that works on largish server architectures to quickly build systems that can work together.

In my spare time i write computer games. Much happier doing this work now.

Answer 7

I'd encourage you to learn about data structures and algorithms. Programming languages come and go, but data structures and algorithms are what is really important.

You've got a PhD in physics, so you're really good at modeling problems. Programming is similar: create a model of the problem, then implement the solution to that problem.

Also, if you use a "Test-Driven Development" methodology when you program, you'll find that programming is very similar to running experiments (a man with a PhD in nuclear physics who is now a programmer told this to me).

Answer 8

Speaking as a programmer with a degree in astrophysics¹, you absolutely can do it. When it comes down to it, the important things to be good at in programming are problem solving and learning (as a general skill). Everything else you can pick up as you go.

I would recommend anything involving 3D graphics or physical simulation, not necessarily games. Having the basic maths knowledge is a huge advantage, even just knowing what a vector or a matrix is.

Good luck!

¹ not many jobs there, either

Answer 9

The answers here seem quite good, so I don't have anything particularly substantial to add, but you're in good company. My friend Marc Fleury is the original author of JBoss and also a physics PhD!

Answer 10

I have worked as a programmer and I am studying for a PhD. I know that McDonald's manager makes more than either. After I graduate, I will go back to business management. There is an oversupply of cheap programmers from all over the world.