More Dask support throughout

[kenkehoe]

Should we refactor our current code to utilize Dask more?

[AdamTheisen]

What do you have in mind? For some of the more heavy compute items, we're already using dask to distribute processing.

io/mpl.py: task.append(dask.delayed(proc_sigma_mplv5_read)(f, save_nc=save_nc,
qc/radiometer_tests.py: task.append(dask.delayed(fft_shading_test_process)(
retrievals/irt.py: task.append(dask.delayed(process_sst_data)(sfc_temp[i], sky_temp[i], emis,
utils/ship_utils.py: task.append(dask.delayed(proc_scog)

[kenkehoe]

I guess we could go through and find more places to add Dask where it is not currently utilized. Are there places where things are slower than they should be?

[AdamTheisen]

Good point and I think there are places we could utilize it but we have to just test. Sometime's it's not faster than just a regular for loop. Just depends on the amount of data and what needs to be done. I think it's worthwhile to investigate though