Open barnzilla opened 6 days ago
Thanks for opening the issue. Just so I can understand more, you don't need to reference a base model run because the compiled model includes all of the input parameters?
Yes, that's correct, Matt. I was chatting with Steve and Anatoly this afternoon and this is a feature of OpenM++ (as opposed to Modgen). A compiled model will have a base scenario in the database (but not a base run).
Okay, then I think it makes sense to make this optional. Let me think a little bit about how to make this change to respect backward compatibility. But generally happy to make this change.
Much appreciated, Matt! And no rush. I have a work-around that I'm using right now.
Sounds good. And thanks for reporting these bugs. I only get to see how openmpp
is used by the users here in our group, so I definitely designed the package to be opinionated for our use cases but I want it to be flexible enough for any OpenM++ user.
Hi @mattwarkentin,
I'm now using
openmpp
in a Linux cloud environment and happy to report that it's working well!One issue I'm having though: I cannot perform a model run with
openmpp
immediately after compiling a model because I do not have a base model run. Currently,openmpp
will abort if the workset object does not have aBaseRunDigest
parameter that is set.Would you be able to either make the
BaseRunDigest
requirement optional or remove it altogether?When I remove the highlighted lines below from
ClassWorkset.R
, I'm able to successfully perform a model run without having a base run:Take care, Joel