[Matplotlib-devel] Overhaul of arrow plotting

Antony Lee antony.lee at berkeley.edu
Wed Mar 8 17:56:10 EST 2017

I obviously have thought much less about this than you, but is there a
reason why you need separate functions to plot in data and in display
space?  Could you not just (similarly to `text`) plot in data space by
default and let the user pass `transform=ax.transAxes` to plot in display
space?  (Yes, there would be some difficulties regarding the arrow head,
which probably always needs to be plotted in display space -- but it should
be possible to take care of that?)

My 2c,


2017-03-08 13:08 GMT-08:00 David Stansby <dstansby at gmail.com>:

> Hi all,
> As I'm sure many of you know, there's currently a bit of a mess of
> functions that can be used for plotting single arrows in Matplotlib. I've
> decided to have a go a tackling the mess with a MEP. I have done a first
> draft here: https://github.com/matplotlib/matplotlib/wiki/MEP-29-(arrows)
> I would appreciate input into defining the end goal; specifically I
> propose that we keep and fix (and possibly rename)
> - matplotlib.patches.YAArrow
> <http://matplotlib.org/api/patches_api.html?highlight=arrow#matplotlib.patches.YAArrow>
> for plotting in 'display space'
> - matplotlib.patches.FancyArrowPatch
> <http://matplotlib.org/api/patches_api.html?highlight=arrow#matplotlib.patches.FancyArrowPatch>
> for plotting in 'data space'
> and deprecate and eventually remove matplotlib.patches.FancyArrow
> <http://matplotlib.org/api/patches_api.html?highlight=arrow#matplotlib.patches.FancyArrow>
> and matplotlib.patches.Arrow
> <http://matplotlib.org/api/patches_api.html?highlight=arrow#matplotlib.patches.Arrow>
> Does this sound sensible? Please send any thoughts/ideas/feedback to the
> mailing list.
> All the best,
> David
> _______________________________________________
> Matplotlib-devel mailing list
> Matplotlib-devel at python.org
> https://mail.python.org/mailman/listinfo/matplotlib-devel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-devel/attachments/20170308/cb2fce09/attachment.html>

More information about the Matplotlib-devel mailing list