hi sir Hernan, thank you soo much for this tutorial it is extremely useful, am Phd student in operational research and i use the same methodology to solve my problem " the graph coloring problem via column generation "but i have a problem in the masterProblem my objectif is to minimize x(s) st: sum x(s) >=1 foreach i in V my question is how I can implement this constrainte without calling the IloNumVar .
Thanks for the tutorials. They're very useful. I have couple of questions regarding modifications of the program: 1- if we are to change the program to only work with one vehicle how we can control the numbers. 2- also can you please give any suggestions to add fuel station to the path if required ? Thanks in advance. Any help would be appreciated.
Thank you for this tutorial it is extremely useful! I am not a OR person, so I have a very basic question. Would you say that this method you are using here is branch and price? If not how is it different? Thank you,
Hi, thank you very much for your tutorial. I am currently trying to convert it to .NET, but I come across a problem. Whenever I apply a vehicle capacity of < 12, no problems occur, but if I increase it to 24, many of the nodes are skipped and the 'zero_reduced_cost_AbortColGen' is reached prematurely. I am not sure what causes this problem and I don't know if anyone can guess it from this limited amount of context. However, if someone could give me a suggestion as to what might be causing this problem, I would be extremely grateful. Thanks again.
thanks for this video, I have a question, I want to write a constraint for master problem like this : 0≤∑y ≤1 instead of ∑y =1 that you used, I am writing it this way: row_taxis.put(taxi, cplex.addRange(0, 1); instead of row_taxis.put(taxi, cplex.addRange(1, 1) but in the model file I am seeing something like this: taxi_0#3: y.(T0)(P0) + y.(T0)(P1) + y.(T0)(P2) - Rgtaxi_0#3 = 0 could you please explain how I should use this method to get the correct output? thanks
Hi Hernan, I am actually having a problem with callbacks in which I'd like to put some lazy constraints in my formulation. Found that you use it in the subproblem of this method but I still can't discover how to use such thing. Any suggestion??
This is a Ph.D student from City University of Hong Kong. Could you please tell me how to generate stochastic scenario-specific costs and material requirements, and related probabilities?
Hi Hernán, I enjoyed to follow you tutorial. But since somebody like me who dont know too much about java, it might be better if you can make another video with easier algorithm with the same problem (VRPTW). Thanks
Hello Hernan, what's up? Well, I'm student of production engineering here in Brazil and I am finishing my graduation. My final paper is about facility location. I have a doubt, I mean, I wanna know if I could include a covering range in a p-median problem. I have the cost matrix (distances between facility and clients), now I wanna put a coverage radius.
Před 10 lety
Yeah, of course, I don't see any problem. Just make sure you model the constraints properly and that's it.
Hernán Cáceres /********************************************* * OPL 12.5 Model * Author: bruno * Creation Date: 04/03/2014 at 16:23:42 *********************************************/ //Data int n = ...; //No de candidatos int m = ...; //No clientes int p = ...; //No de facilidades a localizar range facility = 1..n; //Intervalo do índice das facilidades range client = 1..m; //Intervalo do índice dos clientes //Variables dvar boolean X[facility][client]; //matriz de designação do cliente para cada facilidade dvar boolean Y[facility]; int demand[client] = ...; //matriz de demanda por cliente float dist[facility][client] = ...; // matriz de distâncias entre facilidades e clientes //Objective minimize sum(i in facility, j in client) demand[j]*dist[i][j]*X[i][j]; //Constraints subject to { forall(j in client) sum(i in facility) X[i][j] == 1; // designação de cada facilidade a um único cliente sum(i in facility) Y[i] == p; // garante a localização de p facilidades forall(i in facility, j in client) X[i][j] - Y[i]
I don't know how and where to put the coverage radius...
Před 10 lety
Bruno Oliveira Hey Bruno, I don't know exactly what you mean, there are a few different interpretation that can be derived from "coverage radius". If you give me a more detailed description I might be able to help you. (each facility must cover no more than/at most a certain radius; each client should have a facility within a certain distance; etc.)
thank you soo much, this video saved my life.
I'm a master of mining engineering student of TMU
Its very helpful for me. Thanks.
Hi Hernan, You should totally make Videos with C++ API too
hi sir Hernan, thank you soo much for this tutorial it is extremely useful, am Phd student in operational research and i use the same methodology to solve my problem " the graph coloring problem via column generation "but i have a problem in the masterProblem my objectif is to minimize x(s) st: sum x(s) >=1 foreach i in V my question is how I can implement this constrainte without calling the IloNumVar .
Thanks for the tutorials. They're very useful. I have couple of questions regarding modifications of the program: 1- if we are to change the program to only work with one vehicle how we can control the numbers. 2- also can you please give any suggestions to add fuel station to the path if required ?
Thanks in advance.
Any help would be appreciated.
Thank you for this tutorial it is extremely useful!
I am not a OR person, so I have a very basic question.
Would you say that this method you are using here is
branch and price? If not how is it different?
Thank you,
Hi, thank you very much for your tutorial.
I am currently trying to convert it to .NET, but I come across a problem. Whenever I apply a vehicle capacity of < 12, no problems occur, but if I increase it to 24, many of the nodes are skipped and the 'zero_reduced_cost_AbortColGen' is reached prematurely.
I am not sure what causes this problem and I don't know if anyone can guess it from this limited amount of context. However, if someone could give me a suggestion as to what might be causing this problem, I would be extremely grateful.
Thanks again.
thanks for this video, I have a question, I want to write a constraint for master problem like this : 0≤∑y ≤1 instead of ∑y =1 that you used, I am writing it this way: row_taxis.put(taxi, cplex.addRange(0, 1); instead of row_taxis.put(taxi, cplex.addRange(1, 1) but in the model file I am seeing something like this: taxi_0#3: y.(T0)(P0) + y.(T0)(P1) + y.(T0)(P2) - Rgtaxi_0#3 = 0
could you please explain how I should use this method to get the correct output?
thanks
Is there a way of fixing the number of vehicles here?
Hi Hernan,
I am actually having a problem with callbacks in which I'd like to put some lazy constraints in my formulation.
Found that you use it in the subproblem of this method but I still can't discover how to use such thing.
Any suggestion??
Can we implement CG by using docplex ?
This is a Ph.D student from City University of Hong Kong. Could you please tell me how to generate stochastic scenario-specific costs and material requirements, and related probabilities?
Hi Hernán,
I enjoyed to follow you tutorial.
But since somebody like me who dont know too much about java, it might be better if you can make another video with easier algorithm with the same problem (VRPTW).
Thanks
hello dear, can you record a video lecture on column and constraints generation algorithm?
Could you provide the PDF. It'll help me a lot to learn
dl.dropboxusercontent.com/u/208319/VRPTW.pdf
Thank you a lot.
Hello Hernan, what's up? Well, I'm student of production engineering here in Brazil and I am finishing my graduation. My final paper is about facility location. I have a doubt, I mean, I wanna know if I could include a covering range in a p-median problem. I have the cost matrix (distances between facility and clients), now I wanna put a coverage radius.
Yeah, of course, I don't see any problem. Just make sure you model the constraints properly and that's it.
Is there a problem if I send you my code?, just to take a look, it's small and I have this doubt...
Hernán Cáceres /*********************************************
* OPL 12.5 Model
* Author: bruno
* Creation Date: 04/03/2014 at 16:23:42
*********************************************/
//Data
int n = ...; //No de candidatos
int m = ...; //No clientes
int p = ...; //No de facilidades a localizar
range facility = 1..n; //Intervalo do índice das facilidades
range client = 1..m; //Intervalo do índice dos clientes
//Variables
dvar boolean X[facility][client]; //matriz de designação do cliente para cada facilidade
dvar boolean Y[facility];
int demand[client] = ...; //matriz de demanda por cliente
float dist[facility][client] = ...; // matriz de distâncias entre facilidades e clientes
//Objective
minimize
sum(i in facility, j in client)
demand[j]*dist[i][j]*X[i][j];
//Constraints
subject to
{
forall(j in client)
sum(i in facility)
X[i][j] == 1; // designação de cada facilidade a um único cliente
sum(i in facility)
Y[i] == p; // garante a localização de p facilidades
forall(i in facility, j in client)
X[i][j] - Y[i]
I don't know how and where to put the coverage radius...
Bruno Oliveira Hey Bruno, I don't know exactly what you mean, there are a few different interpretation that can be derived from "coverage radius". If you give me a more detailed description I might be able to help you. (each facility must cover no more than/at most a certain radius; each client should have a facility within a certain distance; etc.)