Implement StokesSpectralCube #249
Conversation
| parameters can be accessed with attribute notation. | ||
| A class to store a spectral cube with multiple Stokes parameters. | ||
|
|
||
| The individual Stokes cubes will share masks. |
|
Looks good; all my comments are minor. I didn't check the travis failure. IO question: what do we do if the user gives a FITS file with 4 dimensions, but one is a degenerate Stokes dimension? I recommend we return a |
|
Right - ideally the user should really access the readers through SpectralCube.read and StokesSpectralCube.read and then it's obvious what the user wants returned. |
|
Good point - yes, let's just have that interface implemented, then. |
|
@astrofrog thanks for the fixes. It looks like there are still bugs. Want to ping me when this is ready for review again? |
|
@keflavich - yes, I'm working on updating the travis config in #258 |
…all yet, and doesn't deal with derived cubes either
…_dir__ with Python 2
|
This still requires some more work for input/output for Stokes cubes, so not ready for review yet. |
|
I don't have time to investigate making a writer for stokes spectral cubes (which requires adding a stokes axis to the WCS), so I'm tidying this up and I think it would be great if we could merge it and address the writing (#266) later. |
|
@keflavich - can you review this? |
| mask[component] = LazyMask(np.isfinite, data=data[component], wcs=wcs_slice) | ||
| comp_data, comp_wcs = cube_utils._orient(data[component], wcs) | ||
| comp_mask = LazyMask(np.isfinite, data=comp_data, wcs=comp_wcs) | ||
| stokes_data[component] = SpectralCube(comp_data, wcs=comp_wcs, mask=comp_mask, meta=meta, header=header) |
There was a problem hiding this comment.
Wrap this at 80 chars? (and related PEP8 if there are other things, but I assume you have done those right everywhere without checking)
|
Review complete! The only comment that requires a chance is the |
|
👍 Merge when it passes |
Implement StokesSpectralCube
@keflavich - this is the combination of your work and my tests and work since.
Still a fair bit to do, including implementing the I/O, but maybe you can already check if this seems ok so far?