-
Notifications
You must be signed in to change notification settings - Fork 121
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.
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
Different result each time problem is solved #488
Comments
Does this happen with any other solver? Does it find different optimal solutions? Or is the problem being solved different each time. Can you provide the logs of different solves? If it's finding equivalent optimal solutions, then I don't really regard that as a bug. I could just be different order of variables/constraints due to dictionary iteration or something. |
If indeed the variables/constraints are getting ordered randomly due to e.g. dictionary iteration, I think to avoid this kind of issue in future, Convex.jl should use e.g. ordered dictionaries. So that the MOI model ends up being exactly the same each time. |
I thought we did use OrderedDicts; I went to double check and saw there is at least one which isn’t: Line 110 in 4f52b6f
Someone could try replacing that (and any others) with an OrderedDict and seeing if that fixes it. |
* Change Dict -> OrderedDict closes #488 * bump version and add release note
This issue was originally opened as jump-dev/Hypatia.jl#801 (comment)
The following non-random problem produces very different results each time the code is run. The problem does not appear to be within the solver Hypatia since if only the part after and including
prob = Convex.minimize(γ)
is run, the results are the same each time.Tested on Convex v0.15.0
The text was updated successfully, but these errors were encountered: