4 ms·

> Traditionally, these sorts of courses at MIT have used Matlab This is an incredibly good sign for hobbyists in a number of fields where Matlab completely dom

by 0x09 13y ago

> Traditionally, these sorts of courses at MIT have used Matlab This is an incredibly good sign for hobbyists in a number of fields where Matlab completely dominates the research end of the spectrum. In the DSP realm the amount of experimental software that relies on Matlab and the number of research publications that pack in Matlab implementations is enough to be a pain in the side for anyone outside of an institution where Matlab licenses are taken for granted. Hopefully it will become a trend.

Bjoern 13y ago
Octave?
- dagw 13y ago
  Octave is close enough to Matlab to have inherited most of its flaws, yet not close enough to be a complete drop-in replacement. Julia, while obviously inspired by Matlab, is not attempting any sort of comparability, and thus can fix flaws and be a better language. Octave, in my opinion, should be relegated to (trying to) run legacy Matlab code, while Julia or Numpy should be the first choice for any new code.
  - jordigh 13y ago
    > Octave, in my opinion, should be relegated to (trying to) run legacy Matlab code, while Julia or Numpy should be the first choice for any new code. Octave dev here. I agree. Every time I have to implement yet another idiocy of the Matlab language, I die a little inside. It's a dirty job, but someone has to do it.
    - papaf 13y ago
      And its greatly appreciated.
    - taliesinb 13y ago
      Come work on a modern functional language!
      - jordigh 13y ago
        You mean mathics? Perhaps. I don't know how important it is to liberate mathematica code. It doesn't seem to have the deployment that Matlab has.
idunning 13y ago
Agreed - its always a sinking feeling to see code attached to a paper and see that its MATLAB or, even worse, MATLAB+some toolkit/addon. Of course, still better than seeing a paper without ANY code...
thearn4 13y ago
Python+Numpy+Scipy+Spyder/IPython have made headway around these parts in displacing MATLAB/Simulink. Simple single installers (Python(x,y), etc) have certainly helped.

by 0x09 13y ago

Bjoern 13y ago

Octave?

dagw 13y ago
Octave is close enough to Matlab to have inherited most of its flaws, yet not close enough to be a complete drop-in replacement. Julia, while obviously inspired by Matlab, is not attempting any sort of comparability, and thus can fix flaws and be a better language. Octave, in my opinion, should be relegated to (trying to) run legacy Matlab code, while Julia or Numpy should be the first choice for any new code.
- jordigh 13y ago
  > Octave, in my opinion, should be relegated to (trying to) run legacy Matlab code, while Julia or Numpy should be the first choice for any new code. Octave dev here. I agree. Every time I have to implement yet another idiocy of the Matlab language, I die a little inside. It's a dirty job, but someone has to do it.
  - papaf 13y ago
    And its greatly appreciated.
  - taliesinb 13y ago
    Come work on a modern functional language!
    - jordigh 13y ago
      You mean mathics? Perhaps. I don't know how important it is to liberate mathematica code. It doesn't seem to have the deployment that Matlab has.

idunning 13y ago

Agreed - its always a sinking feeling to see code attached to a paper and see that its MATLAB or, even worse, MATLAB+some toolkit/addon. Of course, still better than seeing a paper without ANY code...

thearn4 13y ago

Python+Numpy+Scipy+Spyder/IPython have made headway around these parts in displacing MATLAB/Simulink. Simple single installers (Python(x,y), etc) have certainly helped.