New to the python library PULP and I'm finding the documentation somewhat unhelpful, as it does not include examples using lists of variables. I've tried to create an absolutely minimalist example below to illustrate my confusion.

import pulp
IDENTIFIERS = ['A','B','C','D','E']
PRICES      = dict( zip( IDENTIFIERS, [100.0, 99.0, 100.5, 101.5, 200.0 ] ) )
n           = len( IDENTIFIERS )x     = pulp.LpVariable.dicts( "x", indexs = IDENTIFIERS, lowBound=0, upBound=1, cat='Integer', indexStart=[] )
prob  = pulp.LpProblem( "Minimalist example", pulp.LpMaximize )
prob += pulp.lpSum( [ x[i]*PRICES[i] for i in IDENTIFIERS ]  ), " Objective is sum of prices of selected items "
prob += pulp.lpSum( [ x[i] for i in IDENTIFIERS ] )==2, " Constraint is that we choose two items "
prob.solve()
for ident in IDENTIFIERS:if x[ident]==1:print ident + " is in the basket "

The output is:

A is in the basket 
B is in the basket 
C is in the basket 
D is in the basket 
E is in the basket

The optimizer is not recognizing the constraint that we only add two values.

I'll leave this here in case anyone else is just as silly, but actually the above example works fine. I had merely failed to examine the results correctly. Instead:

def printProb( prob ):for v in prob.variables():print v.name, "=", v.varValueprint "Status:", pulp.LpStatus[ prob.status ]

reveals that the solution is correct.