@MarkMcClure thanks, had a quick look. Yes, I see. I am trying to decide now if I should take FEM course or linear systems, so was wondering about how M FEM is used.
but if I take the FEM course, we'll be coding things in Fortran :( (this is what the teacher code is in) but will use M to at least verify my solution. But have not yet decided yet. I like linear systems and control also.
There are a few bugs and limitations and I think that some gui tools to input regions would be nice, especially for beginners, but overall, it's a great tool.
Obviously, I can't advise on FEM vs linear systems.
@MarkMcClure thanks. Yes, all the examples of FEM looks really nice. I think it is easier to use compare to Matlab's FEM toolbox/functions, which I tried and did not like at all.
I actually do like Matlab's PDE toolbox. I'm an intermediate user of Matlab at best and found the GUI to be quite useful. Ultimately, though, the power of the programmatic approach is much more powerful, which is why I think the V10 stuff is such a welcome addition.
Of course, they're both more pleasant than Fortran, I imagine. :)
@MarkMcClure I only used the GUI part of the pde toolbox. Matlab does not have, as far as I know, the regions functions added in V10, very useful. Yes. Fortran is not too easy to use once one is used to M. had to write numbers to a file then load them using M or matlab just to plot something.
I think for control systems, Matlab has more functions and features over M. I was hoping they'll add more stuff to control systems in v10, but only 2-3 functions added, that is all.
@MarkMcClure The reason I think M can have an advantage over Matlab in control (more easily than with numerical) is that one can take advantage of symbolic computation much more with control and dynamics, which M is better at.
Right - I don't know control systems at all but I definitely agree that any numerical scheme that benefits from some sort of symbolic pre-processing is exactly where Mathematica might have an advantage over Matlab.
@MarkMcClure someone should tell Wolfram to hire more control engineer programmers to add more functions to that area. This is an opportunity for WRI to take more advantage of its symbolic powers. There is a book on this subject actually, here it is amazon.ca/Symbolic-Methods-Control-System-Analysis/dp/toc/…