Bugfix/nonlinear pyroscan coordinates

[FelixWattsYork]

There was a bug in my recent PR, which meant that coordinates didn't have the associated values. Basically ky was there but you could only index it by position not value. I have added the proper coordinate adding into the code, and updated the test to check for the correct behaviour

[bpatel2107]

I think this is not preserving the units of the simulation. You might need to add the units in as an attribute

EDIT I mean the units of the coordinates

[bpatel2107]

See these lines here

pyrokinetics/src/pyrokinetics/pyroscan.py

Lines 405 to 407 in 2e23864

	for coord, units in coord_units.items():
	ds[coord] = ds[coord].assign_attrs(units=units)

For older versions of xarray you couldn't assign units directly to the coordinate so you would have to put the unit as an attribute.

However in a newer version now you can so you could force a newer version of xarray where the coordinate has the unit or give it the attribute as above.

[bpatel2107]

@FelixWattsYork I've pushed a temp fix that works for now. Feel free to match the implementation you currently have though and tweak away.

Would be worth adding a test to make sure the units are kept.

@dake0795 checkout this branch please and give it a go?

[FelixWattsYork]

@FelixWattsYork I've pushed a temp fix that works for now. Feel free to match the implementation you currently have though and tweak away.

Would be worth adding a test to make sure the units are kept.

@dake0795 checkout this branch please and give it a go?

Just adding the test now