William Zijie Zhang
William Zijie Zhang
@fanzhuyifan sorry it took some time to review your issue. Does your code work if you don't use the SCIPY backend? i.e. you just try ``problem.solve``? And by last constraint...
> When I briefly ran this model with Mosek I had the impression that in the last constraint (when `i=n-1`) cvxpy generates the sum of everything, ignoring the `axis` argument....
> Should we just add tuple support for bounds instead? I don't see why not.
> Should we add HiGHS as a default solver to support MILPs? yes, I can do it in this PR.
@mkdeak your idea seems quite viable, since having duplicate constraints probably won't affect solve time by much (as they will be disregarded during the pre-solve phase of most solvers). However,...
@mkdeak The warning is only there as a suggestion, if the compile time seems to work for your use-case you can just ignore/suppress the warning. I don't have an answer...
@usamamuneeb might know? Maybe we need to update the cvxpy interface?
> Hello [@cynthia-you](https://github.com/cynthia-you) and [@Transurgeon](https://github.com/Transurgeon)! To the best of my knowledge, SDPT3 is not supported. [#1093](https://github.com/cvxpy/cvxpy/issues/1093) suggests that it was supported in the past (for CVXPY < 1.0). ohh sorry,...
@PTNobel all the CI seems to pass except for the benchmarks (I presume its because no solvers is being installed). Also, could you explain the changes you made to the...
> I am not sure if this is remain of interest. But I want to give `cp.stack` a try. it definitely still is! For ``stack`` I would probably refer to...