Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Verify optimal values reported by ipopt #221

Open
dweindl opened this issue Dec 19, 2019 · 1 comment
Open

Verify optimal values reported by ipopt #221

dweindl opened this issue Dec 19, 2019 · 1 comment
Labels
C++

Comments

@dweindl
Copy link
Member

dweindl commented Dec 19, 2019

[2019-12-19 13:50:47] [INF] [ 0/i02r07c02s12.sng.lrz.de] [o7i4] iter: 4 cost: 163896 time_iter: wall: 367.656s cpu: 4444.31s time_optim: wall: 2062.4s cpu: 22920.6s
[2019-12-19 13:50:47] [INF] [ 0/i02r07c02s12.sng.lrz.de] [o7i5] cachedCost != optimalCost

Add output. Why is a different value reported? Not considering progress during linesearch again?
@dweindl dweindl added the C++ label Dec 19, 2019
dweindl added a commit that referenced this issue May 16, 2020
@dweindl
Handle IpOpt reporting obj_value 0.0 in case of non-finite fval or gr…
8018064 
…adient (#221)
@dweindl
Copy link
Member Author

dweindl commented May 16, 2020

Partially fixed in 8018064.

Remaining cases seem to occur in combination with hierarchical optimization. If Ipopt makes no progress during line search, the optimum is an older function value, for which the inner parameter are not stored anymore. Either need to keep a longer history, or recompute those values.

@dweindl dweindl mentioned this issue Oct 13, 2020
Merged
dweindl added a commit that referenced this issue Oct 14, 2020
@dweindl
Release 0.4.4
0750546 

* AMICI-related
  * Update to AMICI v0.11.2 (#295)
  * adjoint preequilibration
  * reduced memory requirements
  * Make proper use of amici::RDataReporting to avoid computing costly disregarded results (#296)
  * Add script / cmake target to update amici (Closes #288)
  * Handle AMICI failures more gracefully
  * Fix misc/rebuild_amici.sh in case of changed AMICI SWIG interface
* Fix: HierarchicalOptimizationWrapper: Return full parameters and full gradient even if nothing computed analytically
* Fix C++17 in CMakeLists template
* Fix: Use absolute path in misc/setup_amici_model.sh
* Fix PEtab import issues
* Fix icpc option overriding warning (Closes #220)
* Update example notebooks and run on github actions (#294)
* Add CLI option to simulator to run simulations with nominal parameters (#293)
* Various cleanup
  * #287, #285, #271, #270, #268, #267, and others
  * Remove obsolete parpeamici/steadystateSimulator.cpp
  * Remove outdated amiciHelper.py module for parsing AMICI source files
  * Remove all uses of std::tmpnam
* Handle IpOpt reporting obj_value 0.0 in case of non-finite fval or gradient (#221)
* Base image ubuntu:19.10 (EOL ubuntu:19.04)
* adapt dataprovider and simulatio to df to third column in simulation condition table (#272)
* Update gsllite (#269)
* Fix petab import when online some measurements have preequilibrationCondition set
* Add basic contribution guide
* Fix race conditions in steadystate example (#265)
* Cache sonar files on github actions
* CI: No shallow clone for running sonarscanner
* Remove obsolete and unreliabe functions parpe::launchedWithMpi and parpe::initMpiIfNeeded -- using MPI must be specified via CLI now
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
C++
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
1 participant
@dweindl