seems prhmax is our last problematic variable because of cell_method area: mean time: mean within hours time: maximum over hours that should be hardcoded I think, @doblerone ?

It has been decided to delete this variable form the monthly frequency table, only left in the day table.

What is the current behaviour for prhmax @ljoakim ?
Does it correctly read the daily cm_type from the table, i.e. area: mean time: mean within hours time: maximum over hours?

and write?

And does the prhmax cmorization enter this if statement here?
(and what is cm there?)

It has been decided to delete this variable form the monthly frequency table, only left in the day table.

I have removed the monthly prhmax entry from the table
https://github.com/HCLIMcom/HCLIM2CMOR/blob/master/src/CMORlight/Config/CORDEX6_CMOR_HCLIM_variables_table.csv

Here's an example log. I only added the last few lines of the stack trace, and the line number might be a little bit off since it was run a few commits ago:

  File "/home/sm_joalo/dev/repos/HCLIM2CMOR/src/CMORlight/tools.py", line 1447, in process_file
    retval = shell(cmd,logger=logger)
             ^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/sm_joalo/dev/repos/HCLIM2CMOR/src/CMORlight/tools.py", line 44, in shell
    raise Exception('Shell Error: %s for command: %s' % (prc.communicate()[0], cmd) )
Exception: Shell Error: b'\ncdo (Abort): \ndaymean within \x1b[4mhours\x1b[0m time: maximum over hours -seltimestep,1/365 /nobackup/rossby27/proj/rossby/joint_exp/cordex/postprocess/HCLIM2CMOR/data_jl/work/input_CMORlight/EUR11_EUR11_ALADIN43_v1_ERA5_r1i1p1f1_eval/prhmax/prhmax_198501010030-198512312330.nc /nobackup/rossby27/proj/rossby/joint_exp/cordex/postprocess/HCLIM2CMOR/data_jl/temp/1985-a4c92db0-87e0-11ef-8fb8-7cc25546c4cb-prhmax.nc  \n                 ^ Operator cannot be assigned.\n       Reason:\n         No Operators with missing input left.\n\n' for command: cdo -L -f nc4c -s daymean within hours time: maximum over hours -seltimestep,1/365  /nobackup/rossby27/proj/rossby/joint_exp/cordex/postprocess/HCLIM2CMOR/data_jl/work/input_CMORlight/EUR11_EUR11_ALADIN43_v1_ERA5_r1i1p1f1_eval/prhmax/prhmax_198501010030-198512312330.nc /nobackup/rossby27/proj/rossby/joint_exp/cordex/postprocess/HCLIM2CMOR/data_jl/temp/1985-a4c92db0-87e0-11ef-8fb8-7cc25546c4cb-prhmax.nc

So it seems it goes into the elif at line 1440 above, with cm="mean within hours time: maximum over hours", which is part of the total cell method. This causes ncdo to abort. Probably the code a few lines above which extracts cm can not currently handle the complex cell method of prhmax.

Thanks!
OK, so (as expected) it goes into the elif at line 1440 above, with cm="mean within hours time: maximum over hours", and causes cdo to abort.
We don't need to run an additional cdo command here because prhmax is already calculated as daymax in mergemon.sh.
So we need to find a clever way to skip this. A quick and dirty solution would be easy to implement (skip in case var=prhmax) but maybe we can come up with a more general one that handles cm_type=area: mean time: mean within hours time: maximum over hours

I'll have a look at it tomorrow.

Alternatively, we could make a copy of hourly pr in mergemon.sh for the variable prhmax and then make a daymax in line 1440 to get the daily prhmax (analog tasmax etc.). I.e. we try to extract the "max" from the string time: mean within hours time: maximum over hours

and our final solution ? :-)

I didn't have time to implement and test it. It's a nasty variable. I'd think the easiest and most flexible solution is to add another if statement here

This should work both if

1. prhmax is prodced as daymax(pr_1h) in mergemon.sh as it is done now. Then the daymax is simply done twice.
2. prhmax=daymax(pr_ah) is done in master_cmor.sh. Then the prhmax entry in the table https://github.com/HCLIMcom/HCLIM2CMOR/blob/master/src/CMORlight/Config/CORDEX6_CMOR_HCLIM_variables_table.csv needs to start with pr;pr;prhmax and hurly pr needs to be available in input_CMORlight

@ljoakim can you implement and test this? If you don't have the time, I can give it a try.

I'll look into it tomorrow and test it.