Skipping non-existing decision variables in for loop

Answered

Alara Tizer

April 10, 2022 20:28

• 
• 

Hello,

I am trying to maximize the probability of matches p[(i,j)], but because we don't have p[(i,j)] for every z[I,j] value --> the algorithm stops running when it finds an i and j that don't satisfy the p[(i,j)]. Is there a way to skip non-existing variables so the algorithm can continue to run? 

z[I,j]:  the decision variable to have both women i and man j in the group, =1 if both are chosen =, 0 otherwise

Thank you,

0

• Official comment

  

  Simranjit Kaur

  • Gurobi Staff

  November 16, 2025 23:40

  This post is more than three years old. Some information may not be up to date. For current information, please check the Gurobi Documentation or Knowledge Base. If you need more help, please create a new post in the community forum, or try Gurobot, our chatbot interface offering instant, expert-level support.
• 

  Jaromił Najman

  • Gurobi Staff

  April 11, 2022 08:32

  You could define a list or a dictionary of indices that are not present and add an \(\texttt{if}\)-clause to your quicksum call.

  import gurobipy as gp
  
  m = gp.Model()
  
  xI = 1
  xJ = 2
  x  = m.addVars(xI, xJ, name="x")
  yI = 3
  yJ = 4
  y  = m.addVars(yI, yJ, name="y")
  
  present = {}
  # loop over the bigger set yI x yJ
  for i in range(yI):
    for j in range(yJ):
      if i < xI and j < xJ:
        present[i,j] = True
      else:
        present[i,j] = False
  
  m.setObjective(
    gp.quicksum(x[i,j]*y[i,j] 
                for i in range(yI)
                for j in range(yI) 
                if present[i,j]))

  Note that the above snippet works only if \(x_I \times x_J\) is a subset of \(y_I \times y_J\) and might require more work if you are using other indices.

  0

Post is closed for comments.