have QC setup as part of default configuration for all runs?

[eclare108213]

What is the overhead for doing the things needed for the quality control comparisons ("-s qc" to generate and position necessary files) for all simulations? I've done a bunch of simulations without it, and now it would be nice to go back and check the qc output without having to re-run the simulations.

[eclare108213]

A big piece of the "necessary files" is 5 years of daily output. We wouldn't generate that for every run, certainly, but if we were generating it anyhow... maybe a different way to ask this question is this: is there a relatively painless way to retroactively set up what's needed for the qc tests, if we happen to have the output already?

[mattdturner]

If I understand correctly, you are asking about how easy it might be to use the QC script on runs that already generated 5 years worth of daily output? The script should already be setup to properly run with those inputs.

The problem would be that the QC tests (as far as I understand them) require instantaneous output, not averaged output. So, while the script should run with the inputs from any 5 years of daily output, the tests might not be 100% appropriate due to the fact that the default namelist has hist_avg = .true.

[eclare108213]

These are runs that I created for Andrew to analyze, and we kept the aisnap and hisnap fields, which are always instantaneous even when everything else is averaged. I tried running the script on them and I get this error:

$ cice.t-test.py /net/scratch3/eclare/CICE_RUNS/gx1.20yr/history /net/scratch3/eclare/CICE_RUNS/gx1.nonBFB_tmass
INFO:main:Number of files: 0
Traceback (most recent call last):
File "cice.t-test.py", line 441, in
data_a, data_b, data_d, num_files, path_a, fname = read_data(args.base_dir,args.test_dir)
File "cice.t-test.py", line 56, in read_data
nfid = nc.Dataset("{}/{}".format(path_a,files_a[0]),'r')

The script works when I use directories that were generated with -s qc.

[mattdturner]

The script looks for files in the history/ directories that start with iceh_inst.. Since the script says that there are 0 files, I would assume that the files do not start with iceh_inst.. I'd be happy to change the script to look for a filename that might work more universally, I just didn't want the script to also try to read data from files that do not provide the instantaneous output.

[eclare108213]

I see. We'll need to think about this. Options could be to allow hisnap to be read from any file name, or either hi or hisnap from iceh_inst.