AMICI-dev
Adjoint event
Codecov Report
Attention: 58 lines in your changes are missing coverage. Please review.
Comparison is base (
ec4ee25) 77.21% compared to head (7e32da0) 48.57%.
Additional details and impacted files
@@ Coverage Diff @@
## develop #1539 +/- ##
============================================
- Coverage 77.21% 48.57% -28.64%
============================================
Files 94 94
Lines 15089 15149 +60
============================================
- Hits 11651 7359 -4292
- Misses 3438 7790 +4352
| Flag | Coverage Δ | |
|---|---|---|
| cpp | ? |
|
| cpp_python | 37.26% <6.25%> (-0.10%) |
:arrow_down: |
| petab | 53.93% <36.84%> (-0.07%) |
:arrow_down: |
| python | 61.41% <65.78%> (-15.57%) |
:arrow_down: |
| sbmlsuite | ? |
Flags with carried forward coverage won't be shown. Click here to find out more.
| Files | Coverage Δ | |
|---|---|---|
| include/amici/abstract_model.h | 0.00% <ø> (ø) |
|
| include/amici/backwardproblem.h | 0.00% <ø> (-100.00%) |
:arrow_down: |
| include/amici/model.h | 0.00% <ø> (-61.12%) |
:arrow_down: |
| include/amici/model_dae.h | 0.00% <ø> (ø) |
|
| src/abstract_model.cpp | 1.05% <ø> (-5.83%) |
:arrow_down: |
| python/sdist/amici/import_utils.py | 73.52% <66.66%> (-14.71%) |
:arrow_down: |
| include/amici/forwardproblem.h | 0.00% <0.00%> (-100.00%) |
:arrow_down: |
| src/model.cpp | 43.63% <0.00%> (-42.17%) |
:arrow_down: |
| src/model_dae.cpp | 0.00% <0.00%> (-32.10%) |
:arrow_down: |
| src/model_ode.cpp | 46.82% <25.00%> (-20.25%) |
:arrow_down: |
| ... and 3 more |
![codecov[bot] avatar](https://avatars.githubusercontent.com/in/254?v=4 "Published by a pub.dev verified publisher"){kind=small}
SonarCloud Quality Gate failed. 
0 Bugs
0 Vulnerabilities
0 Security Hotspots
7 Code Smells
15.4% Coverage
13.3% Duplication
![sonarqubecloud[bot] avatar](https://avatars.githubusercontent.com/in/12526?v=4 "Published by a pub.dev verified publisher"){kind=small}
SonarCloud Quality Gate failed.
0 Bugs
0 Vulnerabilities
0 Security Hotspots
4 Code Smells
15.5% Coverage
13.4% Duplication
Okay, this seems to be indeed the last status of my work on adjoints with events. Sorry it took me so long to actually double-check it. In principle, the implemented algorithm seems to work fine for multiple test models, but it is - for some reason that I failed to understand - inexact, i.e., gradients are off by a little. This problem only occurred for the Leonhardt test model (purely parameter dependent event with state update). However, this little (in the range of a few percent) being off markedly affected the performance of parameter estimation, as I checked in some (by now) deleted file.
I don't really know what to do with this. The work done here enables adjoint computations for a couple of models with events and heaviside. However, it seems to have a bug for certain types of models with events. Possible reasons might be:
- an error in the implementation
- a problem with interpolation of the forward solution (which is discontinuous), or, as a result, with the right hand side, or dfdp
Okay, this seems to be indeed the last status of my work on adjoints with events. Sorry it took me so long to actually double-check it. In principle, the implemented algorithm seems to work fine for multiple test models, but it is - for some reason that I failed to understand - inexact, i.e., gradients are off by a little. This problem only occurred for the Leonhardt test model (purely parameter dependent event with state update). However, this little (in the range of a few percent) being off markedly affected the performance of parameter estimation, as I checked in some (by now) deleted file.
I don't really know what to do with this. The work done here enables adjoint computations for a couple of models with events and heaviside. However, it seems to have a bug for certain types of models with events. Possible reasons might be:
- an error in the implementation
- a problem with interpolation of the forward solution (which is discontinuous), or, as a result, with the right hand side, or dfdp
when using adjoint sensitivities with events, one should always use polynomial instead of hermite interpolation.
Super late to the party... but could this be the issue of the imprecision?
It seems like Hermite interpolation is used for the forward problem solution, which is also used for the backwards problem.
Super late to the party... but could this be the issue of the imprecision?
It seems like Hermite interpolation is used for the forward problem solution, which is also used for the backwards problem.
Does changing that fix the issue? :)
Super late to the party... but could this be the issue of the imprecision? It seems like Hermite interpolation is used for the forward problem solution, which is also used for the backwards problem.
Does changing that fix the issue? :)
Super late to the party... but could this be the issue of the imprecision? It seems like Hermite interpolation is used for the forward problem solution, which is also used for the backwards problem.
Does changing that fix the issue? :)
Will give it a try... Update: Seem like this didn't fix it.
Quality Gate failed
Failed conditions
0.0% Coverage on New Code (required ≥ 80%)
See analysis details on SonarCloud
Quality Gate passed
The SonarCloud Quality Gate passed, but some issues were introduced.
2 New issues
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code
See analysis details on SonarCloud