Open backeb opened 1 year ago
Great, congratulations!! Note that these are not errors but Warnings. Based on my experience with Delft3D modelling, getting a bunch of warnings is okay :) of course it is wise to check them but only Errors will kill the process. The original model cofigurator, Lauriane, should know most about these warnings. Given the scope of the exercise (running workflow rather than perfet results..) I would ignore them for the time being.
Looking at the map data:
In [2]: import xarray as xr
In [3]: ds = xr.open_dataset('tttz_waq_0007_map.nc')
In [6]: ds.time.data
Out[6]:
array(['2023-03-16T12:00:00.000000000', '2023-03-17T00:00:00.000000000',
'2023-03-17T12:00:00.000000000'], dtype='datetime64[ns]')
Only data until 2023-03-17T12:00:00
is created, but date_min
and date_max
in workflow/config.yml
are 2023-03-16
and 2023-03-18
, respectively. It should have generated data until 2023-03-18T12:00:00
When I did the run yesterday for date_min = 2023-03-15
and date_max = 2023-03-17
, ds.time.data
had more data.
In [21]: ds.time.data
Out[21]:
array(['2023-03-15T12:00:00.000000000', '2023-03-16T00:00:00.000000000',
'2023-03-16T12:00:00.000000000', '2023-03-17T00:00:00.000000000',
'2023-03-17T12:00:00.000000000'], dtype='datetime64[ns]')
Seems like it didn't download cmems
data for 18 March, will need to investigate that.
Running the download rules individually downloads the correct data:
snakemake -R download_cmems_physics --cores 1
and snakemake -R download_cmems_biogeochemistry --cores 1
fetches data up to 18 March.
Will test snakemake --cores 1 all
, to double check things aren't going wrong with the parallelisation.
cc @sebastian-luna-valero
snakemake --cores 1 all
seemed to work without problems, retesting with snakemake --cores 4 all
Now it seems to work...
When running the workflow on
2023-03-31
I get the following errors, but the data is still generated.cc @lorincmeszaros @arthurvd