Passing multiple variables across Python function/methods

Answered

Panggah Prabawa

• August 26, 2024 02:37

Hi all,

I wonder if I can pass multiple/all variables across Python functions considering many variables to pass on so that they are accessible and optimized outside the function. For example:

def vars_constrs(m):
   # Create vars and constraints
   a = m.addVars(range(10))
   b = m.addVars(range(10))
   ......
   z = m.addVars(range(10))
   
   (Create some constraints here)
   ......
   return All_variables

# Main program
m = gp.Model('all')
All_variables = vars_constrs(m)
( Unpack_all_variables(All_variables) )
m.optimize()

# Accessing all variable values
val_a = [a[i].X for i in range(10)]
...
val_z = [z[i].X for i in range(10)]

Is there any way to pack all variables in a function with a single expression All_variables in the function, and then access all the variables when it is called outside the function with Unpack_all_variables? It might be related to m.getvars(), but I am unsure how to do it without indexing.

Thank you.
Best, Panggah.

1

• 

  Ryuta Tamura

  Gurobi Staff

  • August 26, 2024 10:15

  Hi,

  You can pack variables as a List in the function:

  All_variables = [a, b, c]

  Then, variables can be unpacked outside the function like this:

  a, b, c = All_variables

  Is this the behavior you are looking for?

  Thanks,
  Ryuta

  1
• 

  Panggah Prabawa

  • August 27, 2024 01:40

  

  

  

  Hi, Ryuta,

  It can be applicable too, however, I will apply the function iteratively so it may not be easy to write down all the variables that are going to be packed.

  Does Gurobi provide a way to automatically look for existing variables in the model or function before optimizing so that it can be called with just one argument?
  For example, using model.getVars(), I can pack all variables. However, unpacking it will be too tedious because it is a list and I can only access it using an index without the information of the variable names and their tuplelist indices.

  Thank you,
  Best, Panggah.

  0
• 

  Ryuta Tamura

  Gurobi Staff

  • August 27, 2024 02:41

  Hi,

  Unfortunately, Gurobi does not currently provide such a method. Is your goal to have each function take only Model object as arguments and refer to the necessary variable objects in the function? A variable object can be a member of a Model object.

  m._a = a
  m._b = b

  This allows variable objects to be passed to each function along with the model object, so that they can be referenced as needed.

  Thanks,
  Ryuta

  0
• 

  Panggah Prabawa

  • August 27, 2024 06:29

  

  

  

  Hi, Ryuta,

  Sweet. Thank you for your explanation!

  Best, Panggah.

  0

Please sign in to leave a comment.