The parameter values are not updating in the Pyomo model with PySP callback

[Abodh Poudyal]

I believe this is an active group. I have been stuck in this problem for days.

I am using PySP (Pyomo) for one stochastic optimization problem. I created a concrete model for my problem and also defined the scenarios based on the farmer's example given in

https://github.com/Pyomo/pysp/blob/main/examples/farmer/concrete/ReferenceModel.py

In the above example a pysp_instance_creation_callback() function is called for each of the scenarios. In the function, an instance of the model is cloned for each scenario so that the scenario variable (Yield in this case) is updated for each of the scenarios using instance.Yield.store_values(Yield[scenario_name]).

I followed a similar approach to my problem. However, in my case, for each scenario the size of the unknown varies unlike the farmer's example, where the scenarios are for just three crops (wheat, sugar, corn). For instance, my scenarios would look like this,

Scenario1 = {123, 124, 118} 

Scenario2 = {117, 10} 

Scenario3 = {118, 120, 125, 126} 

Scenario4 = {0, 125} ... and so forth

My code snippet looks something like the one below (I have only mentioned useful constraints and variables for simplicity)

# Variable: 

model.nEdges = 129 

model.x_ij = range(0, model.nEdges) # line switching variable range 

model.xij = Var(model.x_ij, bounds=(0, 1), within=Binary) 

  # Scenario parameter: 

model.Fault = Param(mutable=True, initialize={123,124,118}, within=Any) 

# Constraint: 

for key, ite in model.Fault.items(): 

       for faulty in ite.value: 

                  model.c.add(model.xij[faulty] == 0) 

# Scenarios: 

Fault = {} 

Fault['Scenario1'] = {123, 124, 118} 

Fault['Scenario2'] = {120, 124, 118} 

Fault['Scenario3'] = {1, 125} 

# callback function to update the model parameter 

def pysp_instance_creation_callback(scenario_name, node_names): 

             instance = model.clone() 

             instance.Fault.store_values(Fault[scenario_name]) 

             return instance

However, this method did not work for me. The model.Fault value remains the same for each of the scenarios as it was initialized i.e., {123,124,118}. Although if I check the instance value for each of the scenarios, i.e., instance.Fault.value, then it seems like the values are updating (instance.Fault.value gives different values consistent with different scenarios) but while checking the output lp file for the actual model, the constraints are not updated as desired and the final solution comes the same for each of the scenarios as mentioned before. I am not sure how to tackle this issue and I have been stuck in this problem for days. Can anybody help me here?

Regards,

Abodh

[Abodh Poudyal]

I got the solution to this. My instance callback never updated the constraints in the original model. So the workaround would be to fix the variables instead:

model.xij = Var(model.x_ij, bounds=(0, 1), within=Binary) 

# Scenarios: Fault = {} 

Fault['Scenario1'] = {123, 124, 118} 

Fault['Scenario2'] = {120, 124, 118} 

Fault['Scenario3'] = {1, 125} 

# callback function to update the model parameter 

def pysp_instance_creation_callback(scenario_name, node_names): 

                  instance = model.clone() for faulty in Fault[scenario_name]: 

                  instance.xij[faulty].fix(0) 

                  return instance

Thank you  John D. Siirola for the help.