Resolving QPold method of optimizeWBModel #2571
Conversation
Summary:
Failed Tests:No failed tests ✨ Github Test Reporter by CTRF 💚 |
rmtfleming
left a comment
rmtfleming
left a comment
There was a problem hiding this comment.
What is the justification for changing the parameters for how 'ibm_cplex' solves WBMs?
e.g. param.scaind = -1; was put in there for a purpose.
Most WBMs are on the borderline being too multiscale for a double precision solver so the default is to be careful, perhaps slower, to identify a truly optimal solution, than faster but perhaps only approximate the optimal solution.
e.g. test https://opencobra.github.io/cobratoolbox/stable/tutorials/tutorial_numCharactWBM.html with your model
|
My rationale was to make the whole-body model feasible, which worked without setting parameters such as param.scaind = -1. I see why the parameter is necessary now. I was a bit too quick in changing it without understanding the original rationale. Thank you for the clarification and for sharing the link. |
|
I changed solverName to solverNameQP in case QPold as solverName is already defined as character variable in line 145 and this would then result in an error: "Unable to perform assignment because brace indexing is not supported for variables of this type.". |
Summary:
Failed Tests:No failed tests ✨ Github Test Reporter by CTRF 💚 |
|
Farid can this be merged now? |
3 participants
I resolved small errors of the "QPold" option in optimizeWBModel, and added case 'QPold' to the function buildOptProblemFromModel
I hereby confirm that I have:
developas a target branch (top left drop-down menu)