3RiverResearcher
- 合計アクティビティ 11
- 前回のアクティビティ
- メンバー登録日
- フォロー 0ユーザー
- フォロワー 0ユーザー
- 投票 0
- サブスクリプション 3
アクティビティの概要
3RiverResearcherさんの最近のアクティビティ-
3RiverResearcherさんがコメントを作成しました:
Dear Riley,thanks for asking! Gurobi finds an incumbent solution in the Master Problem The Benders Subproblem is solved in the MIPSOL callback An optimality cut is generated because the theta estim...
-
3RiverResearcherさんがコメントを作成しました:
Can I get any feedback on that request? I notice this problem especially on 2 instances. It seems that gurobi adds the cut and immediately forgets that this solution exists.
-
3RiverResearcherさんがコメントを作成しました:
Btw, when I fix the solution to the first cut for the instances where Gurobi does not find any incumbent solution, then it “finds” the solution, but just after two cut loops. This seems a bit weird...
-
3RiverResearcherさんが投稿を作成しました:
Incumbent solution is not updated after optimality cut
ユーザーの入力を待っています。Dear Gurobi staff,I try to understand the behaviour of the Gurobi Solver for a standard Benders Decomposition with lazy callback optimality cuts.I have a multi-period (t) and multi-scenario (s) pro...
-
3RiverResearcherさんがコメントを作成しました:
Dear Gurobi staff, I deleted my post, because I found the error.I thought I had some redundant dual multiplier, which I removed. This was indeed not redundant :DSo please ignore my requests. Have a...
-
3RiverResearcherさんがコメントを作成しました:
Hi Ryuta, thank you very much for your reply! I did not know about the return Value of cbUseSolution. Indeed, if I provide the partial solution I will get the improved objective, if it is the "fu...
-
3RiverResearcherさんがコメントを作成しました:
I noticed one flaw in my approach: The models were run on two different threads, but not in a thread safe way. So I added one distinct gp.Env() to each thread. If I only set a partial solution of x...
-
3RiverResearcherさんが投稿を作成しました:
Super slow "heuristic" solution injection
回答済みDear Gurobi, I have two MIPS: MIP1 and MIP2. MIP1 is a relaxation of MIP2 and uses only a partial subset of all decision variables. But besides that they are the same. I use MIP1 to compute an uppe...