Added min and median combine_functions

[minzastro]

I add two new combine functions. One is simply 'min' - very similar to what exists already.
The other one is 'median', which required some per-block processing. This is probably slower - I made no tests for performance yet.

[codecov]

Codecov Report

Merging #336 (05a9646) into main (185dae1) will increase coverage by 0.38%.
The diff coverage is 98.03%.

❗ Current head 05a9646 differs from pull request most recent head 298424c. Consider uploading reports for the commit 298424c to get more accurate results

@@            Coverage Diff             @@
##             main     #336      +/-   ##
==========================================
+ Coverage   94.27%   94.65%   +0.38%     
==========================================
  Files          24       24              
  Lines         803      842      +39     
==========================================
+ Hits          757      797      +40     
+ Misses         46       45       -1     

Impacted Files	Coverage Δ
reproject/mosaicking/coadd.py	96.19% <98.03%> (+3.76%)	⬆️

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[astrofrog]

Thanks for the pull request! I'll try and take a look at this soon! In the meantime, I noticed that some of the code has been re-formatted - could you run tox -e codestyle to format it back and commit the changes?

[minzastro]

Thanks for the pull request! I'll try and take a look at this soon! In the meantime, I noticed that some of the code has been re-formatted - could you run tox -e codestyle to format it back and commit the changes?

My IDE was complaining that formatting style is non-standard, that's why I changed it. Was this non-standard style intended? I can change it back of course

[astrofrog]

@minzastro - sorry for not getting back to this. Now that we use dask internally in a few places in reproject I think we should also consider using dask to handle the median calculation to avoid having to do all the block handling ourselves. I'll open an alternative PR to try this out.

[astrofrog]

I'm running into issues using dask, have posted a question here: https://dask.discourse.group/t/constructing-a-sparse-dask-array-from-numpy-arrays/2187 - once there is a reply, we can decide how to proceed.