Hi all. Just a little update: I got a few answers from Davide (who did the recent plotting refactors) at https://github.com/mgeier/python-audio/issues/4.
If anyone has further advice, please reply here or there. cheers, Matthias On Tue, Nov 12, 2024 at 9:32 PM Aaron Meurer <[email protected]> wrote: > > On Tue, Nov 12, 2024 at 11:38 AM Matthias Geier > <[email protected]> wrote: > > > > Hi Aaron. > > > > On Mon, Nov 11, 2024 at 8:38 PM Aaron Meurer wrote: > > > > > > I haven't followed all the changes to plotting. Do you know what the > > > PR SymPy was that changed this? > > > > That's not so straightforward ... different things seem to have broken > > at different times. But maybe I wasn't using them correctly in the > > first place? > > > > The example which assigns to "parent" > > (https://nbviewer.org/github/mgeier/python-audio/blob/master/sympy/sympy-matplotlib-animations.ipynb#Changing-the-Whole-Plot) > > still partially works with > > https://github.com/sympy/sympy/commit/5d1e056. > > > > I'm saying "partially", because it now cycles the plotting colors, > > which it didn't do before (see below for the cause of that problem). > > > > One commit later (https://github.com/sympy/sympy/commit/46558492a7), > > I'm getting this exception: > > > > AttributeError: 'MatplotlibBackend' object has no attribute 'fig' > > It seems like that commit was supposed to just be a refactor, so if it > broke some use-case, it was likely unintentional. > > It's not clear to me whether the things you're doing are supposed to > be supported or not (again, I don't really know much about the > plotting module). I would say that animation is something that should > be doable, either by being directly supported or by being > straightforward to hook into what SymPy gives you. > > Aaron Meurer > > > > > The "fig" attribute is now not set in the constructor anymore, but > > instead at a later point, e.g. when calling process_series(). So > > should I call that first? > > > > If yes, it is still not working. Even though this avoids the > > exception, now the animation simply consists of an empty plot (i.e. > > only the axes are visible). > > > > BTW, I'm using "show=False". When I drop that, the "has no attribute > > 'fig'" goes away, but the animation still doesn't work (instead, there > > are a few stray plots generated). > > > > Going back in the history, I found that > > https://github.com/sympy/sympy/commit/79a2cd5432177f662b2bebe1ffc3b3bc4ca575b9 > > is the commit that introduces the color cycling. > > According to the commit message, the color change is definitely > > intentional, so I guess I'll have to try to work around that behavior > > change. > > > > The first example still doesn't work at this point in the history. It > > just shows an empty plot (only the axes are visible). > > > > I wanted to go further back in history, but I wasn't able to install a > > combination of SymPy/NumPy/Matplotlib/Python that worked. > > > > Is there anything else I should try? > > > > cheers, > > Matthias > > > > > Aaron Meurer > > > > > > On Sat, Nov 9, 2024 at 1:24 PM Matthias Geier <[email protected]> > > > wrote: > > > > > > > > Hi all. > > > > > > > > A few years ago, I have documented how to create Matplotlib animations > > > > from SymPy plots: > > > > https://nbviewer.org/github/mgeier/python-audio/blob/master/sympy/sympy-matplotlib-animations.ipynb > > > > > > > > This has worked quite well (as can be seen in the link), but in the > > > > meantime, the behavior/usage of the MatplotlibBackend seems to have > > > > changed and some of the approaches don't work anymore with the latest > > > > SymPy/Matplotlib versions. > > > > > > > > Most notably, the process_series() method doesn't seem to have the > > > > desired effect anymore. > > > > > > > > Note that the approach using get_points() still works perfectly, and > > > > the approach using get_segments() is deprecated, and is not expected > > > > to work anymore. > > > > > > > > Does anyone have an idea how to make the approaches using > > > > MatplotlibBackend work again? > > > > > > > > cheers, > > > > Matthias > > > > > > > > -- > > > > You received this message because you are subscribed to the Google > > > > Groups "sympy" group. > > > > To unsubscribe from this group and stop receiving emails from it, send > > > > an email to [email protected]. > > > > To view this discussion visit > > > > https://groups.google.com/d/msgid/sympy/CAFesC-dx1U1%2BLHDpU4J-SqLs21cszhCaDXSLYSf2dEuTDEjYDg%40mail.gmail.com. > > > > > > -- > > > You received this message because you are subscribed to the Google Groups > > > "sympy" group. > > > To unsubscribe from this group and stop receiving emails from it, send an > > > email to [email protected]. > > > To view this discussion visit > > > https://groups.google.com/d/msgid/sympy/CAKgW%3D6K1hVkWmw3FZ0nHBemS4oBkAqC4kRHdR9uxSXXk6f21EA%40mail.gmail.com. > > > > -- > > You received this message because you are subscribed to the Google Groups > > "sympy" group. > > To unsubscribe from this group and stop receiving emails from it, send an > > email to [email protected]. > > To view this discussion visit > > https://groups.google.com/d/msgid/sympy/CAFesC-cvTU-VOsMgoCRAdWE5jzrSVG5%3Dk-dASH%2BXwaXo8HJ63w%40mail.gmail.com. > > -- > You received this message because you are subscribed to the Google Groups > "sympy" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To view this discussion visit > https://groups.google.com/d/msgid/sympy/CAKgW%3D6KC6QCJt0Y%3DTsa1rFSTRbeMWCgN-WxMxaqx91xQdwb%2BQw%40mail.gmail.com. -- You received this message because you are subscribed to the Google Groups "sympy" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion visit https://groups.google.com/d/msgid/sympy/CAFesC-cWQTsMxQo%3DzigRZZ_rbf59TWyooJ4Q5tp1J4AF5_JYPQ%40mail.gmail.com.
