Use min_ to get element wise minimum of a MVar and 1

Answered

Weihang Zhang

February 21, 2023 22:33

• 
• 
• 

I have the following code and trying to get the elementwise minimum working, however I am getting an error "gurobipy.GurobiError: Invalid data in vars array". Can anyone tell me where I might be doing wrong? 

selection = model.addMVar((1, n1), vtype=gp.GRB.BINARY)

 

product = selection @ condition # condition is a sparse matrix of shape n1xn2

binary = model.addMVar((1, n2), vtype=gp.GRB.BINARY)

for i in range(n):

    model.addConstr(binary[0, i] == gp.min_(product[0, i], 1))

 

Related to

• Matrix API

0

• 

  Riley Clement

  • Gurobi Staff

  February 22, 2023 00:55 Edited

  Hi Weihang,

  The elements in your "product array" are gurobipy.MLinExpr.  The initial arguments to gurobipy.min_ need to be variables.  You can fix this by introducing auxiliary variables like so:

  # could be INTEGER if numbers in "condition matrix" are integer
  product_aux = m.addMVar(n2, vtype=gp.GRB.CONTINUOUS)
  
  for i in range(n2):
      m.addConstr(product_aux[i] == product[0, i])
      m.addConstr(binary[0, i] == gp.min_(product_aux[i] , 1))

  - Riley

  0
• 

  Weihang Zhang

  February 22, 2023 09:51

  • 
  • 
  • 

  Hi Riley,

  Thanks very much for the answer. It is clear now why it was not working previously. Really appreciate the help!

  0

Please sign in to leave a comment.