allow repeat_run from a restart defined in config.yaml #580
Conversation
| @@ -114,7 +114,7 @@ def init_config(self): | |||
|
|
|||
| input_nml = f90nml.read(input_fpath) | |||
|
|
|||
| if ((self.expt.counter == 0 or self.expt.repeat_run) and | |||
| if ((self.expt.counter == 0) and | |||
There was a problem hiding this comment.
I wonder if the original line with or self.expt.repeat_run worked for situations where there isn't a restart: in config.yaml, but we still want to repeatedly start from initial conditions/a cold start? With repeat run, the experiment counter still increments for each subsequent run.
There was a problem hiding this comment.
So it should just be :
if ( self.prior_restart_path is None ):
input_type = 'n'
?
Is prior_restart_path used for both restart paths specified in config.yaml and those generated during payu runs ?
There was a problem hiding this comment.
Yeah, it probably could just be if self.prior_restart_path is None.
Is prior_restart_path used for both restart paths specified in config.yaml and those generated during payu runs ?
Yup, that's correct. The base prior_restart_path is set as part of the Experiment class initialisation to either None, restart directory in config.yaml or an restart in the archive, and then passed to the various model drivers.
Contributes to #576
This enables the repeat option to be used with previous restart files to start (rather than always assuming they are starting from an initial condition).
Ad-hoc testing only. This works with the access-esm1.6 driver now. I've assumed that users specify the restart in
config.yamland do not already have restart's in their archive directory.