Skip to main content

Erreur in my code with gurobipy

Answered
ghada ouerfelli
  • Edited
Gurobi-versary
Conversationalist
First Question

hello I need you help to correct my code 

import gurobipy as gp
from gurobipy import GRB

months = range(5)
products=range(4)
L=100
Dem_client=[[0,0,10,0,0],
[0,20,0,0,0],
[0,0,0,0,15],
[0,10,0,0,25]]

Cout_stockage=[[1,2,3,1,2],
         [6,9,5,1,2],
         [2,1,4,1,3],
         [2,2,1,1,4]]
Cout_chan=[[2,5,8,3,9],
         [1,3,2,4,1],
         [8,9,7,7,4],
         [1,2,1,1,3]]
m=gp.Model('LSP')
x=m.addVars(products,months, lb=0.0)
s=m.addVars(products,months,lb=0.0,obj=Cout_stockage)
y=m.addVars(products,months,obj=Cout_chan,vtype=GRB.BINARY)

m.modelSense = GRB.MINIMIZE
m.addConstrs((x.sum('*',month) <=L)for month in months)
m.addConstrs(s[p][m-1]+x[p][m]==Dem_client[p][m]+s[p][m] for p in products for m,index_m in months if m!=1)
for p in months:
    print (p)
m.optimize()

this is the erreur 

m.addConstrs(s[p][m-1]+x[p][m]==Dem_client[p][m]+s[p][m] for p in products for m, index_m in months if m!=1)
TypeError: cannot unpack non-iterable int object

0

Comments

12 comments

Date Votes
  • Jaromił Najman
    Gurobi Staff Gurobi Staff

    Thank you for providing a reproducible example.

    There are a few issues with your \(\texttt{addConstrs}\) statement. You are using \(\texttt{m}\) as index which simultaneously is the name of your model object. You define \(\texttt{index_m}\), which is not used. Gurobi tupledicts are accessed via standard dict indexing \(\texttt{s[p,m-1]}\) instead of \(\texttt{s[p][m-1]}\). Finally, you check for \(\texttt{m!=1}\) which results in accessing \(\texttt{s[p,-1]}\), which is not defined.

    The following should work in your case

    m.addConstrs((s[p,M-1]+x[p,M] == Dem_client[p][M]+s[p,M] for p in products for M in months if M!=0))

    Best regards, 
    Jaromił

    0
  • ghada ouerfelli
    Gurobi-versary
    Conversationalist
    First Question

    thanks a lot but the best objective is not correct it should be 20 :

    this is the Mathematical formulation:

     

    0
  • Jaromił Najman
    Gurobi Staff Gurobi Staff
    • Edited

    You can use the write method to write a human readable LP file and analyze which constraints are incorrect or missing. You should provide meaningful names to all your variables and constraints before doing so.

    x=m.addVars(products,months, lb=0.0, name="x")
    # [...]
    m.write("myModel.lp")
    m.optimize()

    At a first glance the constraints

    \[x_{p,M} \leq L y_{p,M} \,\, \forall p \in \text{products}, M \in \text{months}\]
    are missing.

     

    Best regards, 
    Jaromił

    0
  • ghada ouerfelli
    Gurobi-versary
    Conversationalist
    First Question

    thanks a lot, Sir but the human-readable LP doesn't work on my Laptop. also, I didn't get the best objective function :

    import gurobipy as gp
    from gurobipy import GRB

    months = range(5)
    products=range(4);
    L=100
    Dem_client=[[0,0,10,0,0],
    [0,20,0,0,0],
    [0,0,0,0,15],
    [0,10,0,0,25]]

    Cout_stockage=[[1,2,3,1,2],
             [6,9,5,1,2],
             [2,1,4,1,3],
             [2,2,1,1,4]]
    Cout_chan=[[2,5,8,3,9],
             [1,3,2,4,1],
             [8,9,7,7,4],
             [1,2,1,1,3]]

    m=gp.Model('LSP')
    #x=m.addVars(products,months, lb=0.0)
    x=m.addVars(products,months, lb=0.0, name="x")
    s=m.addVars(products,months,lb=0.0,obj=Cout_stockage)
    y=m.addVars(products,months,obj=Cout_chan,vtype=GRB.BINARY)
    m.modelSense = GRB.MINIMIZE

    m.addConstrs((x.sum('*',month) <=L)for month in months)
    m.addConstrs(x[p,M]<=L*y[p,M] for p in products for M in months)
    #m.addConstrs(s[p][m]+x[p][m]==Dem_client[p][m]+s[p][m] for p in products for m in months)
    m.addConstrs(s[p,M-1]+x[p,M]==Dem_client[p][M]+s[p,M] for p in products for M in months if M>1)
    m.write("LSP.lp")
    for p in months:
        print (p)
    m.optimize()

    Best objective 1.500000000000e+01, best bound 1.500000000000e+01, gap 0.0000%

    0
  • Jaromił Najman
    Gurobi Staff Gurobi Staff
    • Edited

    Please note that Python indexing starts with \(\texttt{0}\). Thus, the last set of constraints should read

    m.addConstrs((s[p,M-1]+x[p,M] == Dem_client[p][M]+s[p,M] for p in products for M in months if M>=1))

    thanks a lot, Sir but the human-readable LP doesn't work on my Laptop

    What exactly do you mean by "it doesn't work"? Is no file generated? The file should be generated in the execution folder. You could also search your machine for a file named "LSP.lp".

    Best regards, 
    Jaromił

    0
  • Jaromił Najman
    Gurobi Staff Gurobi Staff
    • Edited

    You can let the variable index start at 1 via

    months   = range(1,6)
    products = range(1,5)

    but then you have to adjust the access of the \(\texttt{Dem_client}\) matrix

    m.addConstrs((s[p,M-1]+x[p,M] == Dem_client[p-1][M-1]+s[p,M] for p in products for M in months if M>1))

    You can find more on Python's standard functions and objects at, e.g., W3Schools.

    Best regards, 
    Jaromił

    0
  • ghada ouerfelli
    Gurobi-versary
    Conversationalist
    First Question

    Thanks a lot Sir 

    0
  • ghada ouerfelli
    Gurobi-versary
    Conversationalist
    First Question

    I have another error with lists :  mdl.addConstrs((s[month-1,product]+x[month,product])==(Dem_client[month][product]+s[month,product])
    KeyError: (-1, 0)

    # Parameters and Data
    products=["prod1","prod2","prod3","prod4"] # number of products
    months=["Jan","Fev","Mar","Apr","May"] #number of periode
    Dem_client={
        "Jan":{"prod1":0,"prod2":0,"prod3":0,"prod4":0},
        "Fev":{"prod1":0,"prod2":20,"prod3":0,"prod4":10},
        "Mar":{"prod1":10,"prod2":0,"prod3":0,"prod4":0},
        "Apr":{"prod1":0,"prod2":0,"prod3":0,"prod4":0},
        "May":{"prod1":0,"prod2":0,"prod3":15,"prod4":25}

    }
    Cout_stockage= {
        "Jan":{"prod1":1,"prod2":6,"prod3":2,"prod4":2},
        "Fev":{"prod1":2,"prod2":9,"prod3":1,"prod4":2},
        "Mar":{"prod1":5,"prod2":5,"prod3":4,"prod4":1},
        "Apr":{"prod1":1,"prod2":1,"prod3":1,"prod4":1},
        "May":{"prod1":2,"prod2":2,"prod3":3,"prod4":4}}
    Cout_chan={
        "Jan":{"prod1":2,"prod2":1,"prod3":8,"prod4":1},
        "Fev":{"prod1":5,"prod2":3,"prod3":9,"prod4":2},
        "Mar":{"prod1":8,"prod2":2,"prod3":7,"prod4":1},
        "Apr":{"prod1":3,"prod2":4,"prod3":7,"prod4":1},
        "May":{"prod1":9,"prod2":1,"prod3":4,"prod4":3}}

    L=100# capacité limite de production

    # importing gurobipy package
    from gurobipy import *
    # creating an optimization model
    mdl=Model('LSP')
    # add decision variables
    x=mdl.addVars(months,products,name="x")
    s=mdl.addVars(months,products,obj=Cout_stockage,name="s")
    y=mdl.addVars(months,products,obj=Cout_chan,vtype=GRB.BINARY,name="y")

    #define the objective function
    mdl.modelSense = GRB.MINIMIZE
    #mdl.setObjective(obj, GRB.MINIMIZE)
    mdl.setObjective(0, GRB.MINIMIZE)

    #factory.setObjective(obj, GRB.MAXIMIZE)
    # Add constraints
    mdl.addConstrs((x.sum(month,'*') <=L )for month in months)
    mdl.addConstrs((x[month,product]<=L*y[month,product]) for month in months for product in products)
    mdl.addConstrs((s[month-1,product]+x[month,product])==(Dem_client[month][product]+s[month,product])
    for month in range(len(months)) for product in range(len(products)) if month!=months[0])
    mdl.optimize()
    0
  • Jaromił Najman
    Gurobi Staff Gurobi Staff
    • Edited

    Note that the values of your \(\texttt{months}\) list are now Strings and thus, it is not possible to subtract \(1\) from it. If you want to stick with the String values, you have to redefine your \(\texttt{for}\)-loops in the \(\texttt{addConstrs}\) statement

    mdl.addConstrs((s[previous_month,product]+x[current_month,product])==(Dem_client[current_month][product]+s[current_month,product])
    for previous_month,current_month in zip(months,months[1:]) for product in products)

    Please note that this issue is no longer a Gurobi issue but a pure Python issue and you should refer to a Python forum/tutorial for more insights.

    On a side note, setting of the objective coefficients will not work if you use a "double dictionary". You should either stick with a pure value matrix as in your first post or rewrite the dictionary to hold tuples as keys

    Cout_stockage= {
        ("Jan","prod1"):1,("Jan","prod2"):6,("Jan","prod3"):2,("Jan","prod4"):2,
        ("Fev","prod1"):2,("Fev","prod2"):9,("Fev","prod3"):1,("Fev","prod4"):2,
        ("Mar","prod1"):5,("Mar","prod2"):5,("Mar","prod3"):4,("Mar","prod4"):1,
        ("Apr","prod1"):1,("Apr","prod2"):1,("Apr","prod3"):1,("Apr","prod4"):1,
        ("May","prod1"):2,("May","prod2"):2,("May","prod3"):3,("May","prod4"):4}
    Cout_chan={
        ("Jan","prod1"):2,("Jan","prod2"):1,("Jan","prod3"):8,("Jan","prod4"):1,
        ("Fev","prod1"):5,("Fev","prod2"):3,("Fev","prod3"):9,("Fev","prod4"):2,
        ("Mar","prod1"):8,("Mar","prod2"):2,("Mar","prod3"):7,("Mar","prod4"):1,
        ("Apr","prod1"):3,("Apr","prod2"):4,("Apr","prod3"):7,("Apr","prod4"):1,
        ("May","prod1"):9,("May","prod2"):1,("May","prod3"):4,("May","prod4"):3}

    Best regards, 
    Jaromił

    0
  • ghada ouerfelli
    Gurobi-versary
    Conversationalist
    First Question

    thanks Sir , but the objective result is not correct it is 0: Best objective 0.000000000000e+00, best bound 0.000000000000e+00, gap 0.0000%

    import numpy as np
    import pandas as p
    # Parameters and Data
    products=["prod1","prod2","prod3","prod4"] # number of products
    months=["Jan","Fev","Mar","Apr","May"] #number of periode
    Dem_client={
        "Jan":{"prod1":0,"prod2":0,"prod3":0,"prod4":0},
        "Fev":{"prod1":0,"prod2":20,"prod3":0,"prod4":10},
        "Mar":{"prod1":10,"prod2":0,"prod3":0,"prod4":0},
        "Apr":{"prod1":0,"prod2":0,"prod3":0,"prod4":0},
        "May":{"prod1":0,"prod2":0,"prod3":15,"prod4":25}
    }
    Cout_stockage= {
        ("Jan","prod1"):1,("Jan","prod2"):6,("Jan","prod3"):2,("Jan","prod4"):2,
        ("Fev","prod1"):2,("Fev","prod2"):9,("Fev","prod3"):1,("Fev","prod4"):2,
        ("Mar","prod1"):5,("Mar","prod2"):5,("Mar","prod3"):4,("Mar","prod4"):1,
        ("Apr","prod1"):1,("Apr","prod2"):1,("Apr","prod3"):1,("Apr","prod4"):1,
        ("May","prod1"):2,("May","prod2"):2,("May","prod3"):3,("May","prod4"):4}
    Cout_chan={
        ("Jan","prod1"):2,("Jan","prod2"):1,("Jan","prod3"):8,("Jan","prod4"):1,
        ("Fev","prod1"):5,("Fev","prod2"):3,("Fev","prod3"):9,("Fev","prod4"):2,
        ("Mar","prod1"):8,("Mar","prod2"):2,("Mar","prod3"):7,("Mar","prod4"):1,
        ("Apr","prod1"):3,("Apr","prod2"):4,("Apr","prod3"):7,("Apr","prod4"):1,
        ("May","prod1"):9,("May","prod2"):1,("May","prod3"):4,("May","prod4"):3}

    L=100# capacité limite de production

    # importing gurobipy package
    from gurobipy import *
    # creating an optimization model
    mdl=Model('LSP')
    # add decision variables
    x=mdl.addVars(months,products,name="x")
    s=mdl.addVars(months,products,obj=Cout_stockage,name="s")
    y=mdl.addVars(months,products,obj=Cout_chan,vtype=GRB.BINARY,name="y")

    #define the objective function
    mdl.modelSense = GRB.MINIMIZE
    #mdl.setObjective(obj, GRB.MINIMIZE)
    mdl.setObjective(0, GRB.MINIMIZE)

    #factory.setObjective(obj, GRB.MAXIMIZE)
    # Add constraints
    mdl.addConstrs((x.sum(month,'*') <=L )for month in months)
    mdl.addConstrs((x[month,product]<=L*y[month,product]) for month in months for product in products)
    #mdl.addConstrs((s[months[index_month-1],product]+x[month,product])==(Dem_client[month][product]+s[month,product])
    #for month,index_month in months for product in products if month!=months[0])
    mdl.addConstrs((s[previous_month,product]+x[current_month,product])==(Dem_client[current_month][product]+s[current_month,product])
    for previous_month,current_month in zip(months,months[1:]) for product in products)
    mdl.optimize()
    0
  • Jaromił Najman
    Gurobi Staff Gurobi Staff

    You explicitly set the objective to equal \(0\)

    mdl.setObjective(0, GRB.MINIMIZE)

    Best regards, 
    Jaromił

    0
  • ghada ouerfelli
    Gurobi-versary
    Conversationalist
    First Question

    Thanks a lot, Sir for helping a beginner like me. 

    0

Please sign in to leave a comment.

Didn't find what you were looking for?

New post