Make Mesh.draw a simple wrapper to skfem.visuals.vedo

[kinnala]

Fixes #361.

This is quite cool I think. First pip install vedo. Then try

from skfem import *
m = MeshHex().refined(3)
m.draw(point_data={'y': m.p[0]}).show()

Then pressing 4 and l on the keyboard gives this fancy picture:

Interfacing with vedo is done through a VTK file because the API of vedo is quite complex and this seems to do most of the things that we'd like to do.

[kinnala]

For some reason I can't get this to work in ipython. It detects that I'm in Jupyter Notebook even though I'm not.

[kinnala]

If there are boundaries, then this interactive mode shows only the first one. I wonder what is the easiest way to toggle between those.

[gdmcbain]

I confirm that this works under MS-Windows too, after a pip install vedo. Very nice.

[kinnala]

Trying to do something about boundaries, waiting for marcomusy/vedo#464 to get resolved.

[kinnala]

I mean, it should simply work after the above issue gets resolved, so I guess this can be merged?