orthoCheck Python wrapper is missing its argument

[The9Cat]

It's very trivial. But it's a bug.

The original implementation documented an argument, but the no-argument default was called instead. The fix includes both bindings.

[The9Cat]

• Added additional bug fixes for Eigen::tensor to Python.ndarray conversion
• Give Python user access to the field basis labels
• Fix a mistake in the computation of the polar velocity field

[michael-petersen]

All good fixes. Nice catch on the spherical velocities -- a properly annoying bug!

Our other enhancement debate is whether we want to build an orthocheck into the spherical basis by default before constructing a Cylinder basis. I could go either way on this; maybe something for the future.

[The9Cat]

I think that's a good idea. It's trivial do since the orthoCheck() routines are in both of the SL base classes and there's a global matrix check routine in exputil. Sometime ago, I put in a global orthoTol into the exputil library to allow the user, in principle, to select a orthogonality tolerance. So: one simply needs to put in 'orthoTest` calls which will throw an exception if they don't pass.

[michael-petersen]

Great, this is a fantastic enhancement! We'll have to see how orthoCheck performs in the wild, but my fiducial basis construction worked just fine. I'm in favour of merging this and then sorting out edge cases later if any appear.