Bk/improve error for non integer stoichiometry#1389
Conversation
|
Hello, thanks a lot for the PR. In this case, however, non-integer stoichiometries for bifurcation diagrams are something that we do want to support. |
|
@TorkelE this is only in the context of building the reaction complex mapping though? In that context, it makes sense to have non-integer coefficients, and at some point we should support them in the network analysis as much as possible, but it is true they probably won't really work throughout right now. |
|
sorry, got confussed by reading the title and presumed this was targetting bifurcation kit! |
|
I'm going to reopen this, and will review later today or tomorrow once tests pass. |
|
Seems like tests are failing, so something else needs updating. |
|
One of the test is deprecated, I'm trying to create a fix for it. |
…at, linkageclasses, and deficiency functions
|
Hey @isaacsas, the test worked and now there is no issue with this PR! |
…laws Move incidencematgraph and linkageclasses calls before conservationlaws in deficiency so that the integer stoichiometry check in reactioncomplexmap fires before conservationlaws attempts to compute an integer nullspace on a non-integer stoichiometry matrix. Co-Authored-By: Claude Opus 4.5 <[email protected]>
|
@calebjubal thanks for the PR! All merged now. |
Checklist
contributor guidelines, in particular the SciML Style Guide and
COLPRAC.
Additional context
Added helper function:
Added error check:
Added tests
closes #1286