Taking Statistics vs Pure Math for Experimental Physics

[Emmanouil]

Currently I am studying a double major in Math and Physics, hopefully leading towards experimental physics, however come next semester I will have to pick my math "specialisations" from two of applied math (mostly chaos and optimisation etc.), pure math(mostly abstract algebra and analysis) and statistics/probability (not entirely sure what it would entail). After having spoken to the professors at my school I wasn't really able to find a clear cut answer (some said pure math was vital, while others supported stats, but everyone agreed applied math was extremely important because we have had to cut down on math classes in physics because of the rules of the university). From this I decided that it was probably important to know all 3 before I go much further so I would probably have to pick two to learn from professors and one to learn by myself. I was leading towards getting the credit of stats as I thought it would look better on my resume if I left physics, but I would very much like to know what others further down the line have done/experienced.
Thanks,
Emmanouil

 

[Simon Bridge]

The stats for experimental physics is usually learned by doing ... on the job so to speak.
You don't usually need as much of it as the others.

I'd go for applied and algebra ... algebra has the tools you need, and applied looks like it will help you with modelling.
But there are no right answers here ... that is why you won't get a straight answer. For instance, what sort of experimental physics are you thinking of? Whatever you pick will shape your future work to some extent... but it is usually possible to pick up missed skills later.

 

[chiro]

Hey Emmanouil.

Statistics is the information science (or at least the basis for it) because it deals with information in a sample. Any time you use information it will have some connection in some way to statistics.

Since physics involves information in many ways (statistics, computer work, the fact that the physical universe is composed of information) then on that basis I would recommend statistics over the pure mathematics.

Pure mathematics often doesn't deal with information because the fundamental structures of mathematics (and particularly pure mathematics) don't have a quantization aspect and are often just symbolic. It's not like in computer science where you have to quantize the structure and the algebra and when it comes dealing with information in a specific way, it's often very lacking - even though the concepts are sound.

You aren't going to be held back by your choice for later learning if you don't want to - but I'd say that if it's for learning or reasons of curiosity, the nature of information can be one factor in which to decide whether to choose statistical coursework or pure mathematics coursework (this is aside from choosing one or the other for relevance in regard to your other coursework and ambitions).

 

[Simon Bridge]

Note: The word "quantization" has a special meaning in physics... statistics does not have quantization either.
Compare: "quantification".

 

[chiro]

I meant the term making the information representing something come in chunks with structure.

Real numbers are sort of an exception (unless you have special cases or use symbolic algebra to define them) but statistics is a data/information science since it deals with a finite number of quantities (samples are finite). If you can make the structure of each sample element finite in its memory representation then you can quantize the sample and represent it in a finite number of alphabet letters.

This is not done in pure mathematics since they organize things around arithmetic constraints (like divisibility) and not some quantization of the value itself (i.e. representing a whole number with so many bits).

This can also be one reason why some mathematicians are bad at coding because it forces everything to be absolutely explicit and it's this explicit nature that I think differentiates something like statistics from pure mathematics and to be explicit about information, it has to have representation in a finite number of letters/symbols.