|
Thread Rules
1. This is not a "do my homework for me" thread. If you have specific questions, ask, but don't post an assignment or homework problem and expect an exact solution.
2. No recruiting for your cockamamie projects (you won't replace facebook with 3 dudes you found on the internet and $20)
3. If you can't articulate why a language is bad, don't start slinging shit about it. Just remember that nothing is worse than making CSS IE6 compatible.
4. Use [code] tags to format code blocks. |
On February 03 2017 19:04 Askelad wrote:
Well I don't know any other tool to do scientific computing, except FORTRAN, and I really don't think FORTRAN will solve any of my issues (but maybe I'm wrong). I mean Python or Matlab or whatever are great for small size tests and prototypes but they are way too slow to use with full data :'(
You think you can program something faster than matlab? :O
Scientific computing is a rather broad term, so maybe your specific problem cannot be reduced to a load of matrix operations, but almost every problem I know of can be solved faster in Matlab than in something cobbled together in C++.
Hell, most problems can be solved faster in Python or R, which are significantly less optimized than Matlab, than in something you put together yourself in C++. Especially as they rely on the same libraries (libblas and libatlas) which are written in C and are the most heavily optimized (fee) libraries for linear algebra in existence. They are also relatively bugfree.
Main reason to step away from Matlab I can think of, is to use GPUs. A friend of mine works on GPU algorithms for scientific computing and I think he does use C++. Is that what you're doing?
|
|
|
Yeah, the price is of course quite shocking, but I assumed you have a university license, of your company pays for it. If it's for a hobby project at home, then yeah, the price of Matlab is a very valid concern :p
I don't know much about image segmentation, but everybody I know who works on computer vision swears by Matlab. Especially as images are basically just big matrices.
Matlab is better at loops than it used to be, but it's still not really optimized for them. Generally speaking, you want to write everything you possibly can as a vector (or matrix). It's usually a different way of thinking about the problem, though.
Have you looked at Matlab's optimization toolbox?
|
|
|
There is also octave, a free alternative to Matlab. It is slower but was enough for the computer vision course i took
|
speaking of matlab, I have to use it for my linear algebra class
they don't seem to be interested in teaching us to use it at all, though
guess it will be useful to have some experience with it though
|
On February 04 2017 01:13 travis wrote:
speaking of matlab, I have to use it for my linear algebra class
they don't seem to be interested in teaching us to use it at all, though
guess it will be useful to have some experience with it though
Seems sensible enough. Your linear algebra class is there to teach you linear algebra. Not how to use matlab. A large part of university is (should be) learning how to figure shit out for yourself. Grab a book/tutorial and work through what you need to know. Matlab is pretty easy (as evidenced by all the biologists and physicists who have figured out how to use it  ) for a complex mathematical package.
|
|
|
On February 03 2017 21:19 Askelad wrote:-The fact that for this project I work on 3D images means that most existing image-processing code (Matlab or C++ or else) is not usable (even when the algorithm would not be conceptually different at all!) 
ITK? Combined with VTK / OpenCV that should cover basically all your segmentation / image processing needs.
|
having trouble understanding something in my discrete structures course
question is, for which set D is this true: for all x in D[there exists a y in D ( x < y)]
1.) real numbers
2.) integers
3.) rational numbers
4.) natural numbers
I am confused by the question because isn't the answer just "any of them" ?
Can't you just say that for every x, y = x +1
and do that for real numbers, for integers, for rational numbers, and for natural numbers ?
|
|
|
okay good. it just seemed like a stupid question and I wanted to make sure I wasn't crazy or missing something obvious
|
|
|
Hyrule18969 Posts
integers, rational, and natural numbers are all also real numbers, so the answer is 1
|
On February 05 2017 00:10 tofucake wrote:
integers, rational, and natural numbers are all also real numbers, so the answer is 1
All of them are correct, not just 1. It is true that for all integer numbers, there exists a bigger number.
|
On February 05 2017 04:27 phar wrote:Show nested quote +On February 05 2017 00:10 tofucake wrote:
integers, rational, and natural numbers are all also real numbers, so the answer is 1 All of them are correct, not just 1. It is true that for all integer numbers, there exists a bigger number.
Isn't "real numbers" equivalent to "all of options below" in that question? In that case, I agree with tofucake.
|
Is it asking for one answer or an answer for each of the four sets of numbers? I have a strong dislike for questions where answers are correct, but one of them is more correct than the others.
|
But they're all equally correct in this case?! Tofucake's answer makes no sense.
|
On February 05 2017 10:05 slmw wrote:
But they're all equally correct in this case?! Tofucake's answer makes no sense.
No, they're not. It also depends if you should have one selection only. In that case, you cannot select more than one of them. However, even if you can, real numbers is more correct because it includes irrational numbers. I think irrational numbers still satisfy requirement in that question.
Real numbers = natural numbers + integers + rational numbers + irrational numbers.
Edit: It's a tricky question because you might not think of irrational numbers at all.
|
since it (understandably) turned into a semantics argument, I'll let everyone know that I selected all of the options and got the answer correct.
|
|
|
|
|
|
EPT
