Automatic postprocessing
Contents
Automatic postprocessing¶
Default processing of raw output¶
If all of your models provide a postprocess task called process_raw_output and these tasks should be performed after a run has successfully finished you can add to input/global_settings.py the following lines
################################################
# STEP 8 (optional): Start postprocessing after run has finished #
################################################
process_raw_output = True
This will call the script postprocess.sh of the task for each model in the following way
./postprocess.sh <model-output-dir> <current-start-date> <current-end-date>
where
model-output-dir = “IOW_ESM_ROOT/output/run_name/model”
current-start-date = the start date of the current run
current-end-date = the end date of the current run
Perform specific tasks automatically¶
################################################
# STEP 8 (optional): Start postprocessing after run has finished #
################################################
process_raw_output = False
postprocess_tasks = {
"MOM5_Baltic" : ["process_raw_output", "prepare_for_validator"],
"CCLM_Eurocordex" : ["process_raw_output"]
}
Note that the tasks are preformed in the order they appear in the lists given in the dictionary.
Perform specific tasks with specific arguments automatically¶
################################################
# STEP 8 (optional): Start postprocessing after run has finished #
################################################
process_raw_output = False
postprocess_tasks = {
"MOM5_Baltic" : { "process_raw_output" : {},
"prepare_for_validator" : {"run_name" : "some_other_name",
"init_date" : "19790101",
"end_date" : end_date
}
},
"CCLM_Eurocordex" : ["process_raw_output"]
}
An empty dict means to use the defaults, i.e.
“run_name” = run name as specified in
input/global_settings.py“init_date” = start date of the current run
“end_date” = end date of the current run