Closed JCGoran closed 2 weeks ago
Attention: Patch coverage is 88.00000%
with 3 lines
in your changes missing coverage. Please review.
Project coverage is 85.37%. Comparing base (
5ff221d
) to head (df4a608
).
Files | Patch % | Lines |
---|---|---|
src/visitors/sympy_solver_visitor.cpp | 84.21% | 3 Missing :warning: |
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
Logfiles from GitLab pipeline #217616 (:white_check_mark:) have been uploaded here!
Status and direct links:
It seems that, according to the docs, you can also use SOLVEFOR in KINETIC, which I believe requires modifying some other visitor (which one?): https://github.com/neuronsimulator/nrn/blob/386212697752a52a083b7611a07ab4030d9cb3f0/src/nmodl/simultan.cpp#L372-L389
Logfiles from GitLab pipeline #217825 (:no_entry:) have been uploaded here!
Status and direct links:
The (NON)LINEAR block in NMODL was not accounting for the SOLVEFOR variables, which means that something like the following would fail (when used with
sympy --analytic
):because the SOLVEFOR variables were not accounted for, only the STATE ones, which caused the solver to think this was a nonlinear system due to the presence of the
z * y
term in the second equation (i.e. the solver would try to solve the system for all STATE variables instead of just the ones specified in SOLVEFOR).