sage vs enthought for sci computing

[jadamwilson2]

Hello,
I have recently become interested in using python for scientific
computing, and came across both sage and enthought. I am curious if
anyone can tell me what the differences are between the two, since
there seems to be a lot of overlap (from what I have seen). If my goal
is to replace matlab (we do signal processing and stats on
physiological data, with a lot of visualization), would sage or
enthought get me going quicker? I realize that this is a pretty vague
question, and I can probably accomplish the same with either, but what
would lead me to choose one over the other?
Thanks!

 

[Stef Mientki]

Hello,
I have recently become interested in using python for scientific
computing, and came across both sage and enthought. I am curious if
anyone can tell me what the differences are between the two, since
there seems to be a lot of overlap (from what I have seen). If my goal
is to replace matlab (we do signal processing and stats on
physiological data, with a lot of visualization), would sage or
enthought get me going quicker? I realize that this is a pretty vague
question, and I can probably accomplish the same with either, but what
would lead me to choose one over the other?

Scipy or the new x-y distro is almost identical like MatLab ( but you'll
miss the var browser),
and therefor get started faster.

I've the same kind of applications and support medical researchers with
collecting and analyzing physio data.
About 2 years ago I wasn't satisfied with MatLab anymore,
so I looked at LabView, SciLab, Octave and a few others,
and decided to stay with Matlab, but embed it in my own shell.
Worked with that for about half a year,
then I tried Scipy (also embedded) and found it much better than MatLab.
At the moment I switched totally to Python,
and I'm working on a Labview like environment for Python.

hopes that helps,
cheers,
Stef

 

[three3q]

Hi,

Pylab.
At the moment I switched totally to Python,

Me too, porting scripts was easy.

Dan

 

[sturlamolden]

Hello,
I have recently become interested in using python for scientific
computing, and came across both sage and enthought. I am curious if
anyone can tell me what the differences are between the two, since
there seems to be a lot of overlap (from what I have seen). If my goal
is to replace matlab (we do signal processing and stats on
physiological data, with a lot of visualization), would sage or
enthought get me going quicker? I realize that this is a pretty vague
question, and I can probably accomplish the same with either, but what
would lead me to choose one over the other?
Thanks!

I work in neuroscience, and use Python of signal processing. I've used
Matlab before. Python is just better.

I do not use either Sage or Enthought. Instead I have istalled a
vanilla Python and the libraries I need. The most important parts are:

- Python 2.5.2
- NumPy
- SciPy
- Matplotlib
- wxPython
- pywin32
- PIL
- Cython
- PyOpenGL
- mpi4py
- processing module
- gfortran and gcc (not a Python library, but I need a C and Fortran
compiler)

Less important stuff I also have installed:

- Twisted
- PyGame
- MySQL and mysqldb
- Python for .NET (http://pythonnet.sourceforge.net)
- VideoCapture

 

[Matthieu Brucher]

Hi,

Pylab.

Not a good idea, as pylab will be replaced by pyplot which only
contains plot commands. The remaining will be left to numpy, and pylab
does not have the same functions as numpy.

My answer would be like Stuart's: a mix of Python, numpy, scipy and
whatever scikit or additional numpy module I need, with ipython.
That's if I want to do scientific computation.

Matthieu

 

[Ken Starks]

I work in neuroscience, and use Python of signal processing. I've used
Matlab before. Python is just better.

I do not use either Sage or Enthought. Instead I have istalled a
vanilla Python and the libraries I need. The most important parts are:

- Python 2.5.2
- NumPy
- SciPy
- Matplotlib
- wxPython
- pywin32
- PIL
- Cython
- PyOpenGL
- mpi4py
- processing module
- gfortran and gcc (not a Python library, but I need a C and Fortran
compiler)

Less important stuff I also have installed:

- Twisted
- PyGame
- MySQL and mysqldb
- Python for .NET (http://pythonnet.sourceforge.net)
- VideoCapture

I would add RPy for luck!
The Rproject stats package seems to have attracted a lot of medical
users, and this is a python interface.

I'm not entirely sure what is the advantage of a python wrapper
over R, (compared with the stand-alone Rproject language), but
presumably it would be to combine its functionality with that of
some of the python libraries above.

Anyway, you get lots of graphics for exploratory data analysis, high
quality stats, the ability to write scripts.

The RPy is on sourceforge:
http://rpy.sourceforge.net/

the Rproject itself is at:
http://www.r-project.org/
and there is a whole CRAN (Comprehensive R archive network)

 

[Mike Hansen]

Hello,
I have recently become interested in using python for scientific
computing, and came across both sage and enthought. I am curious if
anyone can tell me what the differences are between the two, since
there seems to be a lot of overlap (from what I have seen). If my goal
is to replace matlab (we do signal processing and stats on
physiological data, with a lot of visualization), would sage or
enthought get me going quicker? I realize that this is a pretty vague
question, and I can probably accomplish the same with either, but what
would lead me to choose one over the other?
Thanks!

Hello,

If you are using Windows, you're better off using Enthought for now
since Sage does not run natively on Windows yet. Hopefully by the end
of the year there will be a native version on Windows.

If you're on Linux or OS X (or Solaris in the near future), then Sage
does have some advantages. If you build Sage from source (which
consists solely of type "make" once), then you get a copy of ATLAS
tuned to your machine which can provide a solid speedup for numerical
linear algebra problems. You also get all of the symbolic and exact
arithmetic which may or may not be beneficial for your application.
Sage also comes with libraries so that you can script Octave or Matlab
or most any other system.

--Mike

Full disclosure: I'm a Sage developer