Hacker News new | past | comments | ask | show | jobs | submit login

> What is it like to understand advanced mathematics?

It's really nice! Get to understand a lot of stuff.

Some of the advanced math is really powerful for applications.

And, can look back at the math saw in physics where wondered if the physics profs really understood the math and conclude, right, they didn't or at least not very well.

E.g., recently I saw a physics lecture where they explained more than once that a quantum mechanics wave function was differentiable and "also continuous" as if there was some question, doubt, choice, or chance otherwise. Of COURSE it is continuous! Every differentiable function is continuous!

There's a lot more on why it's nice to understand the advanced math!




> Every differentiable function is continuous!

Under who's definition? The Heaviside step function is discontinuous, but it's derivative is usually considered to be the Kronecker delta function. Both of these are used extensively in physics and engineering.


Did you mean Dirac delta instead of Kronecker delta?


Yes, my mistake. Kronecker is used often in quantum mechanics, but when working in Hilbert space rather than real space.


> Under who's definition

W. Rudin, Principles of Mathematical Analysis.

W. Fleming, Functions of Several Variables.

For the set of real numbers R, a function

f: R --> R

x in R, d in R, function f is differentiable at x with value d provided for h in R

lim_{h --> 0} (f(x + h) - f(x))/h = d

In this case we write

f'(x) = df(x)/dx = d

In that case, in particular,

lim_{h --> 0} (f(x + h) - f(x)) = 0

so that f is continuous at x.

Differentiation is important in physics, e.g., in Newton's second law.


The Dirac delta function is not a function but a distribution. This is just the product of sloppy notation by the physicists.


No, it is just that physicists means something different than mathematicians when they use the word "function". Just assume that they always are distributions that are evaluated with dirac deltas and everything makes sense.


> If f is differentiable at a point x0, then f must also be continuous at x0.

https://en.wikipedia.org/wiki/Differentiable_function#Differ...


They most likely said "The wave function is differentiable and its derivative is continuous", which isn't obvious. If you think that it sounded strange you should have asked during the lecture and they would have explained. Now instead you just assumed that you knew better and made a fool of yourself...

Also this isn't advanced mathematics, this is something most who are interested learns in high-school and I can assure you that most physics professors are aware of this fact.


> They most likely said

No, they said what I wrote they said. It was at YouTube, from MIT, from a course in quantum mechanics.

Uh, I do understand this stuff: E.g., given a positive integer n, the set of real numbers R, the usual topology for R^n, a set C a subset of R^n closed in that topology, there exists a function

f: R^n --> R

so that for x in R^n, f(x) is 0 for x in C, f(x) > 0 for x not in C, and f infinitely differentiable. E.g., for n = 1, the result applies for set C a Cantor set or a Cantor set of positive measure. For n > 2, the result applies for a sample path of Brownian motion or the Mandelbrot set.

When I was a grad student, I discovered and proved this result. Later I published the result in JOTA.


A bigger problem is the difference between Hermitian and self adjoint, a lot of physics professors don't understand that.

Or for example, this famous dialogue between Geoffrey Chew and Arthur Wightman from the 1960s bootstrap fad:

Wightman asked Chew: why assume from the start that the S-matrix was analytic? Why not try to derive it from simpler principles? Chew replied that "everything in physics is smooth". Wightman asked about smooth functions that aren't analytic. Chew thought a moment and replied that there weren't any.


By the way, the reason they said that was no doubt to motivate the students to choose correct boundary conditions for some Sturm-Liouville problem (as the previous poster sort of alluded to)




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: