#288 also includes some rof fixes that should be implemented when testing this...

Hi @nmizukami , I was able to get these two notebooks to run with cupid-diagnostics -rof --serial with also adding two lines to the ROF additional setup cell (cell 5):

"""
case_dic = {
case_name: {
"grid": grid_name,
"sim_period": slice(f"{start_date}", f"{end_date}"),
"climo_nyrs": min(climo_nyears, int(end_date[:4]) - int(start_date[:4]) + 1),
"output_dir": CESM_output_dir,
}
}
if base_case_name is not None:
case_dic[base_case_name] = {
"grid": grid_name,
"sim_period": slice(f"{base_start_date}", f"{base_end_date}"),
"climo_nyrs": min(
climo_nyears, int(base_end_date[:4]) - int(base_start_date[:4]) + 1
),
"output_dir": base_case_output_dir
}
"""
It looks like obs_available must have gotten set to false, though, since no obs plots showed up.

I'm not sure why the non-serial method isn't working for me. Did you run this with dask successfully?

Hi @nmizukami , I was able to get these two notebooks to run with cupid-diagnostics -rof --serial with also adding two lines to the ROF additional setup cell (cell 5):

""" case_dic = { case_name: { "grid": grid_name, "sim_period": slice(f"{start_date}", f"{end_date}"), "climo_nyrs": min(climo_nyears, int(end_date[:4]) - int(start_date[:4]) + 1), "output_dir": CESM_output_dir, } } if base_case_name is not None: case_dic[base_case_name] = { "grid": grid_name, "sim_period": slice(f"{base_start_date}", f"{base_end_date}"), "climo_nyrs": min( climo_nyears, int(base_end_date[:4]) - int(base_start_date[:4]) + 1 ), "output_dir": base_case_output_dir } """ It looks like obs_available must have gotten set to false, though, since no obs plots showed up.

I'm not sure why the non-serial method isn't working for me. Did you run this with dask successfully?

Hi Teagan, I believe I was able to run with dask. I executed cupid-diagnostics -rof . But, I did see a bunch of message from a dask, which I thought could be ignored. I can re-run and post what that says.

Yes, obs_available should be False.

Hi @TeaganKing , I put a log from cupid-diagnostics -rof here.
/glade/work/mizukami/CUPiD/examples/key_metrics/cupid-diagnostics-rof.log.local.

I am wondering about cupid-infrastructure. I have not update this env for long time. wondering if you have updated some packages?

Ok, thanks Naoki. It seems like it is a sporadic error with dask-- eg, this case had a failure:
/glade/work/tking/cupid_project/other_cupids/234v238/CUPiD/examples/key_metrics/

but this one was fine:
/glade/work/tking/cupid_project/other_cupids/234v236.002/CUPiD/examples/key_metrics

We should figure out what is causing the occasional dask errors, as they would seem concerning to users even if the plots are generated successfully.