Unusual behavior when using Pyomo and Pandas (?) #68
Description
Hi all,
I'm getting an extremely unusual behavior when using Pyomo with a Pandas Series object. Specifically, when I convert a Pandas Series object to a dictionary and pass it to a Param initialization, I get a error when I create a constraint rule.
import pandas as pd
from pyomo.environ import *
model = ConcreteModel()
model.Set1 = Set(initialize=[0, 1, 2, 3, 4, 5])
# model.Parameter1 = Param(model.Set1, initialize=pd.Series({0: 400.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0, 5: 240.0}).to_dict())
model.Parameter1 = Param(model.Set1, initialize={0: 400.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0, 5: 240.0})
model.Variable1 = Var(model.Set1, initialize=0)
def rule(m, l):
return -m.Parameter1[l] <= m.Variable1[l]
model.Constraint1 = Constraint(model.Set1, rule=rule)
The above works fine. However, if I replace the initialize argument for Parameter1 with a Series object converted to a dictionary, I get the following error.
ERROR: Constructing component 'Constraint1' from data=None failed:
ValueError: Constraint 'Constraint1[0]' does not have a proper value. Found 'True'
Expecting a tuple or equation. Examples:
summation(model.costs) == model.income
(0, model.price[item], 50)
Traceback (most recent call last):
File "test.py", line 21, in <module>
model.Constraint1 = Constraint(model.Set1, rule=rule)
File "/Users/$USER/anaconda/lib/python2.7/site-packages/pyomo/core/base/block.py", line 483, in __setattr__
self.add_component(name, val)
File "/Users/$USER/anaconda/lib/python2.7/site-packages/pyomo/core/base/block.py", line 849, in add_component
val.construct(data)
File "/Users/$USER/anaconda/lib/python2.7/site-packages/pyomo/core/base/constraint.py", line 752, in construct
cdata = self._check_skip_add(ndx, tmp)
File "/Users/$USER/anaconda/lib/python2.7/site-packages/pyomo/core/base/constraint.py", line 900, in _check_skip_add
condata.set_value(expr)
File "/Users/$USER/anaconda/lib/python2.7/site-packages/pyomo/core/base/constraint.py", line 442, in set_value
raise ValueError(msg)
ValueError: Constraint 'Constraint1[0]' does not have a proper value. Found 'True'
Expecting a tuple or equation. Examples:
summation(model.costs) == model.income
(0, model.price[item], 50)
In the example above, I've commented out the line that converts a dictionary to a Series object and back to a dictionary. To reproduce this error, uncomment that line and comment the following line.
I'm using Pyomo 4.4.1 and Pandas 0.17.1
I've also verified that the two arguments are equivalent.
In [1]: import pandas as pd
In [2]: {0: 400.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0, 5: 240.0} == pd.Series({0: 400.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0, 5: 240.0}).to_dict()
Out[2]: True
Any idea what is going on? Most of the data I have is in Pandas Dataframes / Series. It is puzzling to me why this does not work.