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
Assertion `model->col_upper_[col] == kHighsInf || (model->col_upper_[col] >= implColUpper[col] - primal_feastol && colUpperSource[col] == row)' failed. #1688
Comments
Interesting, could you provide the MPS file? |
I note that it says "options file not found", so if you're wanting to reproduce options settings that you make in your application, it's not picking them up. I didn't look at your MPS file, but there's a limit to the precision of real values in MPS |
No option file is used in either the call from the C++ API or from console. The path is set, so that I could change the options of the subsolver (e.g. HiGHS ) without recompilation.
Is there an option to generate a MPS file with extended precision (Xpress used to need a flag to save in full precision). I attached the MPS file below, but as I said, from the console I can't reproduce the bug either. |
I see that your MPS file has plenty of full-length values. The option you describe would be relatively easy to implement. I'm puzzled as to why that "Options file not found" is logged |
I think this is because I do not check if the file exists, but always forward the path to HiGHS // try to read settings from file
model.readOptions(this->settings_file); |
Generating it with more significant digits seems to be easy indeed. (I just replaced "10g" with "17g" in HMPSIO.cpp). |
I tried the following to debug/reproduce the issue:
BTW: |
Unfortunately, I am not able to reproduce this yet. I suspect that its due to the fact that the problematic assertions do not take into account that the implied column bounds may be infinite, i.e. I have tried to fix this in my fork: https://github.com/fwesselm/HiGHS/tree/fixAssertionFailure @Downsite, could you try this out? Thank you. |
Thanks! |
Similar to #1684, it appears when solving the problem from C++, but I can't reproduce it from the MPS file. Maybe there is either a difference between the command line solver and the C++ API, or the MPS file is not an exact representation of the state of the optimizer?
Anyways, the problem is as follows:
The text was updated successfully, but these errors were encountered: