Potential issue with CGYRO signs for theta grid

[physycola]

I have a case comparing GENE, GS2 and CGYRO for the same linear mode (input_files.zip) where the sign of the eigenfunctions is opposite to the other two codes, as you can see below (CGYRO is in red). The mode has a non-zero theta0, but it doesn't appear to be an issue with the theta0-related logic in cgyro.py. Any ideas?

[physycola]

Note that this Also appears to extend to the sign-convention associated with the fluxes as well. I haven't been able to track down exactly where the change needs to be made yet.