Error with deepcopy (pickle) applied to ConcreteModel

[Cord Kaldemeyer]

Hi everybody,

when using deepcopy on a Concretemodel (which I need for a specific problem), I get an error:

TypeError: can't pickle dict_keys objects

Here's a modified example from the repository that illustrates the problem:

#  ___________________________________________________________________________
#
#  Pyomo: Python Optimization Modeling Objects
#  Copyright 2017 National Technology and Engineering Solutions of Sandia, LLC
#  Under the terms of Contract DE-NA0003525 with National Technology and
#  Engineering Solutions of Sandia, LLC, the U.S. Government retains certain
#  rights in this software.
#  This software is distributed under the 3-clause BSD License.
#  ___________________________________________________________________________

#
# Knapsack Problem
#

from copy import deepcopy
from pyomo.environ import *

v = {'hammer':8, 'wrench':3, 'screwdriver':6, 'towel':11}
w = {'hammer':5, 'wrench':7, 'screwdriver':4, 'towel':3}

limit = 14

M = ConcreteModel()

M.ITEMS = Set(initialize=v.keys())

M.x = Var(M.ITEMS, within=Binary)

M.value = Objective(expr=sum(v[i]*M.x[i] for i in M.ITEMS), sense=maximize)

M.weight = Constraint(expr=sum(w[i]*M.x[i] for i in M.ITEMS) <= limit)

# check
M2 = deepcopy(M)

Dict-keys in this case are of type string and I am wondering why it is not working.

Any hints on how I can create a deep copy of a ConcreteModel?

Thanks in advance!

Cheers
Cord

[William Hart]

Can you give more details?  The compute platform?  Version of python and pyomo? Etc

--
You received this message because you are subscribed to the Google Groups "Pyomo Forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pyomo-forum+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Cord Kaldemeyer]

Hi Will,

of course! I haven't thought about having the latest version and now updated from Pyomo 4.4 to Pyomo 5.2 which gives me another error message:

ERROR: Unable to clone Pyomo component attribute.
    Component 'ITEMS' contains an uncopyable field 'initialize' (<class 'dict_keys'>)

I am on Ubuntu 16.04 LTS x64 with Anaconda 3.5 and Pyomo 5.2 installed via conda-forge:

pyomo                     5.2                      py35_0    conda-forge

What's wrong about the dict that's used to initialize the set?

Cheers
Cord

To unsubscribe from this group and stop receiving emails from it, send an email to pyomo-forum...@googlegroups.com.

[Siirola, John D]

This is a Python 3.x problem.  “dict.keys()” returns an unclonable dict_keys object instead of a list.  We try and convert simple things like tuples and generators into lists, but it turns out that dict_keys is not a GeneratorType (<sigh>).

 

My simple fix to more aggressively convert any iterable thing into a list caused several test failures, so we will have to think more about how to properly fix it (filing an issue in GitHub would be helpful).

 

In the meantime, as a workaround for Python 3.x, do not initialize a Set with a dict.keys(), and instead explicitly convert it to a list: “list(dict.keys())”.

 

john

 

 

From: pyomo...@googlegroups.com [mailto:pyomo...@googlegroups.com] On Behalf Of Cord Kaldemeyer
Sent: Monday, January 08, 2018 6:09 AM
To: Pyomo Forum <pyomo...@googlegroups.com>
Subject: [EXTERNAL] Re: Error with deepcopy (pickle) applied to ConcreteModel

 

Hi Bill,

of course! I haven't thought about having the latest version and now updated from Pyomo 4.4 to Pyomo 5.2 which gives me another error message:

ERROR: Unable to clone Pyomo component attribute.
    Component 'ITEMS' contains an uncopyable field 'initialize' (<class 'dict_keys'>)

I am on Ubuntu 16.04 LTS x64 with Anaconda 3.5 and Pyomo 5.2 installed via conda-forge:

pyomo                     5.2                      py35_0    conda-forge

Thanks.

[Siirola, John D]

FYI: PR 305 (https://github.com/Pyomo/pyomo/pull/305) has a proposed fix for this bug.

 

john

[Cord Kaldemeyer]

Hi John,

thanks for your answer! How are the chances to merge the PR into the next release?

Otherwise, I'll have to rewrite my code with you hotfix which would also be o.k..

Best
Cord

[Cord Kaldemeyer]

I'll just keep track on github.

Thanks!

[Siirola, John D]

I can guarantee the fix will be in the next release. I am actually hoping to have the PR into master by the end of this week. 

John

[Cord Kaldemeyer]

Thanks a lot!

Best
Cord