Brilliant! Thank you Rob! ... View more

That is interesting to know about the results being different on different machines but makes sense as long as the conditions are met.  I'm still a bit confused on what I have wrong with my constraints because the solver is providing results that are greater than my customer orders. For example, CustomerB ordered 10 of Product B but the solver is giving them 161.   Rob I have to again thank you for holding my hand on this, it's clear I do not fully understand this procedure but I'm planning on reading several of your papers this weekend. ... View more

Thanks again for your effort on this, would you mind posting your full optmodel code? I'm getting the same objective value but not the same matrix of customers/products. data orders_costs; length customer $25. product $25. revenue 8. num_orders 8.; input customer & product & revenue num_orders; cards; CustomerA ProductA 300 100 CustomerA ProductB 200 130 CustomerB ProductA 300 0 CustomerB ProductB 200 10 CustomerC ProductA 300 10 CustomerC ProductB 200 60 ; data supply; length product $25. supply 8; input product & supply; cards; ProductA 85 ProductB 180 ; proc optmodel; /* declare sets and data indexed by sets */ set <str> O; set <str> D; number a{O}; number b{D}; number c{O,D}; number n{O,D}; /* read data */ read data supply into O=[product] a=supply; /*supply*/ read data orders_costs into D=[customer] b=num_orders; /*customer demand*/ read data orders_costs into [product customer] c=revenue n=num_orders; /*customer-product demand*/ /* solve variable and model description */ /*matrix of supply and demand*/ var x{O,D} >= 0; /*model objective*/ max total_revenue = sum{i in O, j in D}c[i,j]*x[i,j]; /*constraints*/ constraint supply{i in O}: sum{j in D}x[i,j]=a[i]; constraint demand{i in O, j in D}: x[i,j] >= 0.1*n[i,j]; /* solve and output */ solve; print x; /* create data */ create data sparse from [customer product] x; create data dense from [product]=O {j in D} <col(j)=x[product,j]>; quit; Output:    ... View more

Thanks Rob -- I'm not sure if the updated demand constraint is working correctly. I'm expecting at least 10% of a customer's orders to be filled up until the supply constraint runs out.   For example, in my data CustomerA has placed 100 orders for ProductA and CustomerC has placed 10 orders. Total supply for ProductA is 85, and demand is 110.   The solver outputs the below result which is strange since in this simple example the revenue is the same for this product I would expect to see something like 84 units for CustomerA and 1 unit for CustomerC. data orders_costs; length customer $25. product $25. revenue 8. num_orders 8.; input customer & product & revenue num_orders; cards; CustomerA ProductA 300 100 CustomerA ProductB 200 130 CustomerB ProductA 300 0 CustomerB ProductB 200 10 CustomerC ProductA 300 10 CustomerC ProductB 200 60 ; data supply; length product $25. supply 8; input product & supply; cards; ProductA 85 ProductB 180 ; data demand; length product $25. demand 8; input product & demand; cards; ProductA 110 ProductB 200 ; proc optmodel; /* declare sets and data indexed by sets */ set <str> O; set <str> D; number a{O}; number b{D}; number c{O,D}; number n{O,D}; /* read data */ read data supply into O=[product] a=supply; /*supply*/ read data orders_costs into D=[customer] b=num_orders; /*customer demand*/ read data orders_costs into [product customer] c=revenue n=num_orders; /*customer-product demand*/ /* solve variable and model description */ /*matrix of supply and demand*/ var x{O,D} >= 0; /*model objective*/ max total_revenue = sum{i in O, j in D}c[i,j]*x[i,j]; /*constraints*/ constraint supply{i in O}: sum{j in D}x[i,j]=a[i]; constraint demand{j in D}: sum{i in O}x[i,j] >= 0.1*b[j]; /* solve and output */ solve; print x; /* create data */ create data sparse from [customer product] x; create data dense from [product]=O {j in D} <col(j)=x[product,j]>; quit;   ... View more

Once again, you're the best! Thank you Rob! ... View more

Rob -- I hate to ask again but my needs have slightly changed. In the below example I'm trying to solve a problem where I have more customer orders than I can fill for different product lines. I'd like to be able to maximize revenue and fill at least 10% of each customer's orders (where possible). I must not be understanding the array aspect of reading data into the optmodel procedure, what do I have incorrect below? Thanks @RobPratt!   data orders_costs; length customer $25. product $25. revenue 8. num_orders 8.; input customer & product & revenue num_orders; cards; CustomerA ProductA 300 100 CustomerA ProductB 200 130 CustomerB ProductA 300 0 CustomerB ProductB 200 10 CustomerC ProductA 300 10 CustomerC ProductB 200 60 ; data supply; length product $25. supply 8; input product & supply; cards; ProductA 85 ProductB 180 ; data demand; length product $25. demand 8; input product & demand; cards; ProductA 110 ProductB 200 ; /*Note: optimum_orders_to_fill is hypothetical*/ data want; length customer $25. product $25. optimum_orders_to_fill 8. num_orders 8.; input customer & product & optimum_orders_to_fill num_orders; cards; CustomerA ProductA 84 100 CustomerA ProductB 120 130 CustomerB ProductA 0 0 CustomerB ProductB 5 10 CustomerC ProductA 1 10 CustomerC ProductB 55 60 ; proc optmodel; /* declare sets and data indexed by sets */ set <str> O; set <str> D; number a{O}; number b{D}; number c{O,D}; number n{O,D}; /* read data */ read data supply into O=[product] a=supply; /*supply*/ read data orders_costs into D=[customer] b=num_orders; /*customer product demand*/ read data orders_costs into [customer product] c=revenue; /*revenue*/ read data orders_costs into [customer product] n=num_orders; /*customer demand*/ /* solve variable and model description */ /*matrix of supply and demand*/ var x{O,D} >= 0; /*model objective*/ max total_revenue = sum{i in O, j in D}c[i,j]*x[i,j]; /*constraints*/ constraint supply{i in O}: sum{j in D}x[i,j]=a[i]; constraint demand{j in D}: sum{i in O}x[i,j]=b[j]; /*ability to add constraint where each customer has at least 10% of their orders filled?*/ /* solve and output */ solve; print x; /* create data */ create data sparse from [customer product] x; create data dense from [product]=O {j in D} <col(j)=x[product,j]>; quit; ... View more

This is brilliant Rob, thank you SO MUCH! You are what makes the SAS community great! ... View more

Wow, thank you so much Rob. I just translated the example into 3 datasets -- here they are:   data costs; length origination $25. destination $25. cost 8; input origination & destination & cost; cards; Detroit Boston 30 Detroit New York 20 Pittsburgh Boston 40 Pittsburgh New York 10 ; data supply; length origination $25. supply 8; input origination & supply; cards; Detroit 200 Pittsburgh 100 ; data demand; length destination $25. demand 8; input destination & demand; cards; Boston 150 New York 150 ; ... View more

Thanks! ... View more

Thanks Cynthia! In my case, I didn't need to use the tableeditor tagset, so I switched to ODS HTML and it worked!   Thanks! ... View more