Pass custom built MPL colormaps

[banesullivan]

Feature request: allow passing custom made Matplotlib colormaps

import matplotlib.pyplot as plt
cmap = plt.cm.get_cmap("viridis", 5)

then pass that colormap to the Viewer

(the goal isn't necessarily catecorical colormaps... this is just a simple example. A user might want to make a custom normalized map that are far more complex)

[thewtex]

Support is added for images in #213

[quaeritis]

I'm trying to do this for point_set and it doesn't seem to work, both viewer = view(point_sets=None, annotations=False, cmap=itkwidgets.cm.bone) and viewer.cmap = 'gist_earth' have no effect.

So how can I set the colormap for point_set?
It would also be very helpful to be able to set the opacity for point_set.

[thewtex]

@quaeritis This can currently be achieved with, e.g.:

view(point_sets=[point_set_1, point_set_2, point_set_3],
     point_set_opacities=[0.3, 0.8, 0.4],
     point_set_colors=[[1.0, 0.1, 0.0], [1.0, 0.3, 0.0], [1.0, 0.7, 0.0]])

In this notebook:

https://mybinder.org/v2/gh/InsightSoftwareConsortium/itkwidgets/master?filepath=examples/NumPyArrayPointSet.ipynb

[quaeritis]

Hi, thanks for the fast reply. But I would like to use the colormap to color the scalar values.
It should look like this in the plot on the right:

[thewtex]

👍 we are working towards this.