AttributeError: 'gurobipy.Var' object has no attribute 'X' when model is optimized

回答済み

Jack Heavey

2023年11月14日 19:56

I am creating a Gurobi model to solve a minimization problem across a series of graphs, minimizing the distance from a node across a sample of graphs.

I have confirmed that the model is optimized, outputting model.Status of 2 and model.SolCount of 1. However, when trying to access the individual variable attributes using the var.X attribute, I get the title attribute error. Within prior debugging, I have seen that the .X attribute does not exist if the model has not yet been optimized, but mine has. Is there another way that I can get these values or fix this error? Thank you!

関連

0

コメント

4件のコメント

Riley Clement

Gurobi Staff

2023年11月14日 23:12
Hi Jack,

It will be hard to help without seeing some code. Are you able to share your code via a share drive, or github, or something like Google Colab?

- Riley
0
Jack Heavey

2023年11月14日 23:21
Hi Riley,

The relevant code and files can all be found in the below Github repository:

https://github.com/Jack-Heavey-12/pool_testing/tree/main

The files 'pool.py' and 'lambda_pool.py' are both having the same errors.

Thank you for your help!
0
Riley Clement

Gurobi Staff

2023年11月15日 01:06
Hi Jack,

I'd guess the issue is that you're creating the model within a function, and after the function is complete Python's garbage collector is removing the model, and consequently solution values.

A couple of solutions:

1) return the model object from the function, and assign it, so that it is not removed, or

2 you can assign the X values to an attribute on the variables, eg
```
def LinearProgram(graph, set_list, cascades, B=3, overlapping=True):
    ...
    variables = m.getVars()
    for v in variables:
        v._X = v.X
    return x, y, m.objVal, variables

...

def rounding(x_dict):
   ...
        if limit <= x[S]._X:
            ...
```
- Riley
0
Jack Heavey

2023年11月15日 19:33
Thank you Riley, this did resolve the issue.
0

サインインしてコメントを残してください。