Closed Doresic closed 6 months ago
Attention: Patch coverage is 85.71429%
with 2 lines
in your changes are missing coverage. Please review.
Project coverage is 84.26%. Comparing base (
9cf4bf2
) to head (476c2b6
).
Files | Patch % | Lines |
---|---|---|
pypesto/visualize/model_fit.py | 50.00% | 2 Missing :warning: |
:exclamation: Your organization needs to install the Codecov GitHub app to enable full functionality.
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
There was a bad implementation of inner parameter returning on call of objective. Previously, the
InnerCalculatorCollector
would take note ofself.best_fval
such that it returnsinner_parameters
only in cases a better objective function value was found. This was ok during optimization, but then if one wanted to get inner parameters for some other outer parameter vector for which the objective function would not be better, the inner parameters would not be returned.This is bad practice: to save current states in the calculator and objective objects. So now, after an optimization start is finished, the
hierarchical_decorator
instead calls the objective once more with the best outer parameters found to obtain the inner parameters for that outer vector.There some small update in the visualize of
model_fit.time_trajectory_model
. Was not working with the current implementation due to recent hierarchical structure changes.