Re: [matplotlib-devel] ginput in nbagg backend to use in IPython Notebooks

[Mark B. <ma...@gm...>]

Thanks, Tom.

I want to use ginput to draw a straight line on a graph.
The line is used to select a cross-section of a contour plot.

I was afraid it wasn't going to be easy.

Getting to it from the other side, is there a matplotlib widget in the
works where I can type text or numbers in a box? Like the FloatTextWidget
in IPython?

Problem is I want to make a small GUI that includes both a text widget
(which is available in IPython) and a 'select points in graph' widget like
ginput in matplotlib.

Mark


On Mon, Jan 26, 2015 at 11:47 PM, Thomas Caswell <tca...@gm...> wrote:

> nbagg is always running in the IPython event loop (as I understand it), so
> I am not sure how to integrate that with the blocking.
>
> On the 1.4.x/master branch we have support for (almost, one PR still
> pending) all mouse and keyboard events so all of the mpl widgets should
> work (big thanks to Steven Silvester).  T
>
> What do you want to use that relies on ginput?
>
> You can fake up a non-blocking version something like:
>
> from collections import deque
> ```
> class accumulator(object):
>     def __init__(self, n=5):
>         self.list_of_points = deque(maxlen=n)
>
>     def on_event(self, event):
>         self.list_of_points.append(event)
>
> import matplotlib
> import itertools
> import numpy as np
> matplotlib.use('nbagg')
> import matplotlib.pyplot as plt
> plt.close('all')
> fig, ax = plt.subplots()
> x = np.linspace(0,10,10000)
> y = np.sin(x)
> ln, = ax.plot(x,y)
>
> dd = accumulator(15)
> fig.canvas.mpl_connect('button_press_event', dd.on_event)
> plt.show()
> ```
>
> and then get the points by
>
> ```
> dd.lest_of_points
> ```
>
> This code obviously needs lots of bells and whistles, but points in the
> right direction.
>
> Tom
>
> On Mon Jan 26 2015 at 2:45:45 PM Mark Bakker <ma...@gm...> wrote:
>
>> Hello List,
>>
>> Are there any plans to make ginput work in the nbagg backend?
>>
>> It would be so cool if I could use that in an IPython Notebook together
>> with the other widgets.
>>
>> Thanks,
>>
>> Mark
>> ------------------------------------------------------------
>> ------------------
>>
>