Update FITS-cubes (in quarantine)

[eblur]

The FITS-cubes tutorial is currently in the quarantine branch.

Issues that need to be addressed:

• cube.quicklook() raised errors with matplotlib. May need updates to make it compatible with spectral-cube version 0.4.3. The file was made with version 0.4.1

• We need to find a file that is less than a 200 MB download.

[eblur]

Opened an issue in spectral-cube repo:
radio-astro-tools/spectral-cube#511

[eblur]

@kakirastern is working on this

[kakirastern]

Thanks @eblur. Am looking into the issue, and will open a PR regarding it soon.

[kakirastern]

The FITS-cubes tutorial is currently in the quarantine branch.

Issues that need to be addressed:

• cube.quicklook() raised errors with matplotlib. May need updates to make it compatible with spectral-cube version 0.4.3. The file was made with version 0.4.1

Response: Matplotlib is currently in version 3.0.2, while Spectral-Cube in version 0.4.4. Will strive hard to update code (or maybe potentially a rewrite) to render it compatible with the latest version of both packages, plus potentially others. (Current setup: matplotlib=2.1.0, spectral-cube=0.4.1)

• We need to find a file that is less than a 200 MB download.

Response: I have some data cubes for some planetary nebulae that are much smaller than the 200-MB limit; but if I use these this means that a new tutorial will be needed to be written, as the subject matter at hand would no longer be "a simple visual analysis of the Small Magellanic Cloud (SMC) using HI 21cm emission and a Herschel 250 micron map", so I don't recommend using my samples. Alternatively, I can try to either reduce the current data cube size by "data binning" or by looking for another suitable one by scouring the Internet for some good ones.

Somewhere down the line I may be able to contribute a second tutorial using my own data and experience. It is my strong believe that the current work should be updated, rather than repurposed.

[kakirastern]

I have found the directory containing the original data cube used in the tutorial, at the URL: http://cdsarc.u-strasbg.fr/ftp/J/A+A/594/A116/CUBES/GAL/TAN/, and have checked that file sizes listed there are all greater than 200 MB.

[kakirastern]

So finding a smaller replacement file from the same source is not an option. Next, will attempt to reduce chosen data cube used sample in file size with rebinning, perhaps using SciPy... (Example: https://scipy-cookbook.readthedocs.io/items/Rebinning.html).

[kakirastern]

Issues that need to be addressed:

• cube.quicklook() raised errors with matplotlib. May need updates to make it compatible with spectral-cube version 0.4.3. The file was made with version 0.4.1
• We need to find a file that is less than a 200 MB download.

Further responses:

• cube.quicklook() seems to be working fine with Matplotlib v3.0.2 and Spectral-Cube v0.4.3 now... so will move on the other parts of tutorial.
• It takes more than 10 minutes to download the original cube, so alternative route of downloading it first then read it from computer would also not be recommended.

[kakirastern]

Opened a PR to address this Issue.

[kakirastern]

So the Google CoLab link for the updated version of the tutorial can be found at https://colab.research.google.com/drive/1dX05QJ0aOyVJN9ZWWP2-2CCnneW1ew8o .