search

SemBioProcess / SemGen

A tool for semantics-based annotation and composition of biosimulation models
18 stars 3 forks source link

Variables with annotations in the groups of "encapsulation" are not detected by SemGen for auto-mapping while merging. #329

Closed Niloofar-Sh closed 4 years ago

Niloofar-Sh commented 4 years ago

While trying to merge a model (CellML format) with the resultant of 2 previously merged models, SemGen cannot detect and suggest the same annotated variables for auto-mapping which exist in the encapsulated groups.

maxneal commented 4 years ago

Hi Niloofar - could you provide some more details about this issue? I'm not sure I completely understand the problem. Is it that you have a 2-step merging task and variables that you need map in the second step are being hidden because of something that happens in the first step? If you could attach the models that you are using in the merge as well as the resulting model that would help, too.

Niloofar-Sh commented 4 years ago

Hi Max! Yes, that's exactly the case. Here, first the "aortic_root_module" is merged with "Ascending_aorta_module". I annotated some of the variables in them and for instance I indicated the source/sink for a number of fluid pressures (u & u_in). The result of the first merge (here is named "block2") is going to be merged with another model (here is "brachiocephalic_module"). So what happens next is that the variables which were previously annotated in the brick modules of "block2", don't pop up again as equivalent variables with the ones in "brachiocephalic_module" (supposedly because of the encapsulation which has happened while merging the first two modules). Also, it seems that in "block2", the added sources/sinks have disappeared from the previous annotation. Maybe I'm missing something while merging the first two.🤔

On Thu, 9 Jul 2020 at 04:29, maxneal notifications@github.com wrote:

Hi Niloofar - could you provide some more details about this issue? I'm not sure I completely understand the problem. Is it that you have a 2-step merging task and variables that you need map in the second step are being hidden because of something that happens in the first step? If you could attach the models that you are using in the merge as well as the resulting model that would help, too.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/SemBioProcess/SemGen/issues/329#issuecomment-655624109, or unsubscribe https://github.com/notifications/unsubscribe-auth/AMNFTY5OZJAWNGE2ZZ4SU6DR2SNEXANCNFSM4OTVFNBQ .

maxneal commented 4 years ago

I suspect SemGen didn't preserve some of the annotations correctly during the first merging step. Could you email me the files that you are trying to merge so I can reproduce the issue on my end?

Niloofar-Sh commented 4 years ago

Yes, sure! The 2 PNG files show how the modules are supposed to be connected (merged). And also, "Final_merged_SemGen.cellml" is the result of all these merges. The file was run and the results were validated.

[image: Slide5.png] [image: Slide4.png]

On Fri, 10 Jul 2020 at 12:19, maxneal notifications@github.com wrote:

I suspect SemGen didn't preserve some of the annotations correctly during the first merging step. Could you email me the files that you are trying to merge so I can reproduce the issue on my end?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/SemBioProcess/SemGen/issues/329#issuecomment-656413213, or unsubscribe https://github.com/notifications/unsubscribe-auth/AMNFTY72DIOM7O34YZRKQC3R2ZNATANCNFSM4OTVFNBQ .

maxneal commented 4 years ago

Fixed in commit 9423279dd2ef0e824ac5239dcc2d79b407bab161. The issue was not due to encapsulation, but rather the incorrect assignment of generic metadata IDs to CustomEnergyDifferential instances.