| Re: Any interest in an IDL to Python interface? [message #59061 is a reply to message #59008] |
Mon, 03 March 2008 01:19  |
![Anthony[1] is currently offline](theme/default/images/xoffline.png.pagespeed.ic.XRkd1fkXye.png "Anthony[1] is currently offline")
Anthony[1]
Messages: 20
Registered: December 2006
|
Junior Member |
|
|
On Mar 2, 6:57 pm, Michael Aye <kmichael....@googlemail.com> wrote:
> On Mar 1, 3:49 am, metachronist <rkombi...@gmail.com> wrote:
>
>
>
>> On Feb 29, 9:42 am, Jason Ferrara <jason.ferr...@jacquette.com> wrote:
>
>>> On Feb 28, 12:13 pm, Reimar Bauer <R.Ba...@fz-juelich.de> wrote:
>
>>>> Hi
>
>>>> it could be nice if something like that is added for all supported OS
>>>> versions not only for windows.
>
>>> Its our plan to support Windows, Linux, and Mac.
>
>>>> python has lots of interesting libraries and is used as scripting
>>>> language in many programs (OpenOffice.org, Blender, Maya, PyMOL, Gimp, etc.)
>
>>>> The problem will be the price or do you have considered to share it free
>>>> of charge?
>
>>> It would be a commercial product. The pricing hasn't been determined
>>> yet.
>
>>>> And another question has to be discussed too. What of idl will be left
>>>> over if such a powerful programming language will be added to idl.
>>>> For example this will give us the possibility to use QT4 for widgets. Or
>>>> we never again do get "the sky is falling down miracle" because of the
>>>> decimal data type. Or we can use pythons standalone webserver or lots of
>>>> math libraries free of charge.
>
>>>> One has to ask himself if an idl program is mostly based on python
>>>> modules why he has to use idl and does not do the whole job in python
>>>> then? There are not much differences between both languages.
>
>>>> How would the development of idl been continued if we as user could
>>>> always use other libs?
>
>>> I don't see this as being an issue.
>
>>> Each language has its own strengths and weaknesses. Its not a one
>>> size
>>> fits all thing. Interoperability between languages means you can mix
>>> and
>>> match as best solves your problem, rather than having to pick one and
>>> then struggle with it for the things its not good with.
>
>>> And more libraries accessible from IDL makes IDL more useful, not
>>> less.
>
>> First of all this is an excellent idea.. But just curious.. what is
>> the memory overhead in computationally intensive apps? Any bench
>> marks? I personally am a big python advocate and learning more
>> everyday! Thanks for any additional info that you can provide.
>> Best wishes,
>> /rk
>
> Just adding:
> Me too, plz. I love Python and would at least like to have some
> interoperability.
> I'm also eagerly waiting for a Python-SPICE interface, then SPICE
> results could be fed directly into IDL's powerful plotting routines.
> Or IDL's results could be fed via Python into this Python-supported 3D
> rendering program Blender.
> Go ahead with it!
> Best regards,
> Michael
Great idea - definitely worth having access to both Python and IDL
functionality at the same time.
For myself, I think I'd generally want to use IDL within Python,
rather than Python within IDL - though it would be good to be able to
do both.
There are a couple of ways of running IDL from Python. E.g.,
pyIDL <http://www.its.caltech.edu/~mmckerns/software.html> - looks
great, but I couldn't get it to compile on my machine.
pIDLy <http://astronomy.sussex.ac.uk/~anthonys/pidly/> or <http://
pypi.python.org/pypi/pIDLy/> - a module I wrote as a wrapper on
Pexpect. More likely to compile, but pretty slow for large arrays
etc., not likely to get much faster, and at an early stage of
development.
Cheers,
Anthony
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
