[IPython-user] Workspaces with Embedded IPython

[Fernando Perez]

Oops, forgot to CC the list...

On Thu, Aug 14, 2008 at 2:36 PM, Fernando Perez <fperez.net@gmail.com> wrote:
> On Thu, Aug 14, 2008 at 6:09 AM, Domain Admin <adowd@dspthunder.com> wrote:
>> Hello Fernando and Stefan -
>>
>> Thanks for the information.  If the branch created by Gaël is the
>> direction the ipython-wxpython integration is taking, I'm certainly
>> interested in following an existing trail.  I'll start perusing  the
>> specified code branch.  Thanks.
>
> You can see this code in the trunk now, it's been all merged.
>
>> Please allow me to bore everyone with an outline of my project (and
>> indirectly respond to Ryan's question).  Everyone's help on the
>> workspace issue was so useful, I'd love to some general guidance
>> before I waste too much time going down blind alleys.  Some of my
>> design issues are beyond the scope of this ipython e-mail list.   I
>> plan to start posting to pertinent bulletin boards, but I'd rather
>> list all my main architecture questions;  just ignore points that are
>> irrelevant.
>>
>>  My goal is to create an OS (BSD-license) tool for processing
>> time&frequency data.  This is a fairly small niche, but one that would
>> be well-served by the growing python-based replacements for Matlab.
>> It requires specialized algorithms (I'm 75% done writing these),
>> decent plottings capability (2D, Log-log and some 3D), some
>> interactions with plotting and customizable scripting (hence the need
>> for workspace interaction).  Using python, ipython, numpy and wxpython
>> as the foundation seems pretty certain.  I'm still evaluating the
>> various plotting engines (matplotlib, chaco).
>
> Sounds great!  I can only encourage you to look at Gael's work and the
> Enthought tools, in particular Envisage (even if you use matplotlib
> for plotting).  The wx shell was written *precisely* with integration
> with the rest of the Enthought toolchain and GUIs in mind, something
> we all want.  If you need 3d views, that system also gives you
> mayavi2.
>
> I'm sure you'll find plenty of interested users on the numpy/scipy
> lists as well as here (though the ipython readership contains users
> from many backgrounds outside of scientific computing).  In what
> pertains to ipython, we'll be happy to help where we can.  More
> importantly, let us know how the current design works or fails to work
> for you: we *want* the new codebase to be easy to use for this type of
> thing, so anything that's a problem for you will be considered a bug
> by us.
>
> Cheers,
>
> f
>