tag:blogger.com,1999:blog-593563533834706486.post573559598479887279..comments2024-03-28T10:35:10.453-04:00Comments on Yet Another Math Programming Consultant: 2d Bin PackingErwin Kalvelagenhttp://www.blogger.com/profile/09496091402502236997noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-593563533834706486.post-9204304138527958062021-02-17T07:53:10.317-05:002021-02-17T07:53:10.317-05:00You can simultaneously strengthen and shrink Model...You can simultaneously strengthen and shrink Model D by replacing <= 1 with u[k] and then omitting the x <= u constraints, which are then dominated.Rob Pratthttps://www.blogger.com/profile/16525877394541155854noreply@blogger.comtag:blogger.com,1999:blog-593563533834706486.post-71201051008656012942021-02-16T15:17:35.386-05:002021-02-16T15:17:35.386-05:00I think in my case I need x(item,row,col,bin). Abo...I think in my case I need x(item,row,col,bin). About 1.2e6 variables if I did not make a mistake. Worth a try...Erwin Kalvelagenhttps://www.blogger.com/profile/09496091402502236997noreply@blogger.comtag:blogger.com,1999:blog-593563533834706486.post-53679381470154823042021-02-16T13:56:34.971-05:002021-02-16T13:56:34.971-05:00I looked at a similar model many years ago. In my ... I looked at a similar model many years ago. In my case, everything fell onto a grid, so I wrote a very different model: x[i,j,k] if item k is in position i,j. Then I had a constraint for every position in the grid: sum (ii,jj,k covering i,j) x[i,j,k] <= 1 for all grid positions (i,j). Performance was much better - no disjunctive constraint.Greg Glocknerhttps://www.blogger.com/profile/14964879286706089383noreply@blogger.com