You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm using ECOS (Python) as a subroutine to solve Exponential Cone Programs. I have used it for two different projects and in both, I noticed that up 3% of the instances fail to be solved and the reason is Combined backtracking failed 0 0 90 0 sigma 1
As a far as I understood from "Ecos" Thesis that problem lies in the line search.
The question is there something which I can do within ECOS to tune the line searching method?
I'm using ECOS (Python) as a subroutine to solve Exponential Cone Programs. I have used it for two different projects and in both, I noticed that up 3% of the instances fail to be solved and the reason is
Combined backtracking failed 0 0 90 0 sigma 1
As a far as I understood from "Ecos" Thesis that problem lies in the line search.
The question is there something which I can do within ECOS to tune the line searching method?
ECOS output for the instance reported is
P.S. I have other similar instances. If anyone needs to test the instances, I can share the GitHub repository and explain how to use it.
Cheers
The text was updated successfully, but these errors were encountered: