search

maxwnewcomer / NWTOPT

NWTOPT is a hyperparameter optimization and tuning system for steady-state and transient models ran by MODFLOW-NWT
Creative Commons Zero v1.0 Universal
2 stars 0 forks source link

set IPRNWT to consistent value #18

Closed hwreeves-USGS closed 3 years ago

hwreeves-USGS commented 3 years ago

The IPRNWT parameter just sets printing of iteration information so it probably should be consistently set to 0 or 1.

Setting it to 0 gives a slightly faster overall time because MODFLOW-NWT doesn't have to write out all the iteration information; but its nice to have it set to 1 for checking model performance. Maybe the time difference between the same set of parameters with IPRNWT 0 or 1 is not that important; but it should be something.

I suggest setting it to 0 for NWTOPT, users can easily flip it to 1 if they want to look at where the model is getting bogged down. You are not saving the list files in NWTOPT, so setting the flag to 1 would not be helpful in the optimization.

HWR

maxwnewcomer commented 3 years ago

The only thing I can think of where setting it to 1 would be useful is if you are checking your runs through the logs/output/ directory during a HTCondor run. I do agree that it should just be a consistent value though.

hwreeves-USGS commented 3 years ago

If you set it to 1 the information is printed to the lst file - so the user can't see it in the output log. The output log is capturing output to the screen. The user can ask HTCondor to return the lst files from every run; but that seems like overkill for NWTOPT. If it seems useful down the road we could write a simple script and related sub file that would send out the top X, or some random number, of nwts to workers, edited so they all have the flag set to 1, and have the lst files returned to condor_logs. That could be useful if some model is having trouble and the user wants to compare iteration behavior for several nwts. Much lower priority than MODFLOW6, GSFLOW, etc.