Re: gnucap-python

On Sun, Apr 8, 2018 at 3:25 PM, Felix Salfelder <[email protected]> wrote:

> On Mon, Feb 19, 2018 at 12:26:06AM +0100, Felix Salfelder wrote:
> > On Sun, Feb 18, 2018 at 12:27:02PM +0100, Patrick Mulder wrote:
> > > gnucap> python loadplot.py
> >
> > can't find any loadplot.py. i assume that loadplot.py actually implements
> > the "myplot" command.
> >
> > > ImportError: numpy.core.multiarray failed to import
> >
> > still don't understand. to me it seems that this exception is thrown
> > before anything interesting happens.
>
> Hi Patrick & all
>
> digging a bit deeper, i could figure out some of the things that Henrik
> tried to do.
>
> I now have working example sessions using components and simulation
> commands written in python. one other python script combines nlopt and
> gnucap, implementing some basic parameter optimization.
>
> The numpy related code, and the warnings are still in. i don't
> understand what this was supposed to achieve. fancier stuff has to wait
> for the output rewrite...
>
> if anybody is interested: swig supports more than just python [1], it
> would be fun to turn the package into gnucap-swig, with support for more
> than one and then grow with swig.
>

Hi Felix,

thanks, this sounds great!
I will try the Python update soon. I also looked a bit more into Swig and
it looks interesting to learn more on it. The original problem why we
started looking into Python though might be interesting to recall:
The Python libraries for plotting data are very interesting. The Python
exentension could make it easier to plot:
* i-v curves of devices
* transient simulation of e.g. logic circuits
* AC frequency sweeps

While circuits and simulations can be easily defined with Gnucap, the
plotting is something I want to refresh for myself for some basic
circuits...

Nice day,

Patrick