Objective function in Python

[Phuc Nguyen]

Hi,

How can I write my objective function in Python language using predefined sets and parameter values? The param values are stored in matrices. The python program returns no value for me.

Ampl code:

from amplpy import AMPL, Environment, DataFrame

import numpy as np

ampl = AMPL(Environment(r'D:\Hello Dream 2020\FSS 2021\OPM 682\5. Software\ampl_mswin64'))

ampl.eval('option version;')

ampl.setOption('solver', 'knitro')

ampl.eval(r'''

# ---- CAOP / bilinear mixed 0-1 formulation ---- #

#### Define sets ####

set M;                      # Set of products

set N;                      # Set of customer segments

set K;                      # Set of resources that may constrain the assortment

#### Define parameters ####

param gamma{M};             # Probability that the demand originates from segment i

param rho{M,N};             # Unit revenue of product j from segment i

param v{M,N};               # Preference of product j to segment i

param v0{M};                # Preference of the no-purchase option to segment i

param Beta{K,N};            # Amount of resource k used by product j

param Capacity_Max := 5;    # Maximum capacity of the assortment

#### Define variables ####

var x{N} binary;            # Decision identifier to include product j into the assortment (=1 for yes) and (=0 for no)

var y{i in M} >= 0;

#### Define objective function ####

maximize Revenue: sum{i in M, j in N} gamma[i] * rho[i,j] * v[i,j] * y[i] * x[j];

#### Constraints ####

subject to Capacitated_Capacity {k in K}: sum{j in N} Beta[k,j] * x[j] <= Capacity_Max;

subject to MMNL_model_nonlinear {i in M}: y[i] = 1 / (v0[i] + sum{j in N} v[i,j] * x[j]);

##### Define problem #####

problem CAOP_Bilinear_Mixed_0_1: x, y, Revenue, Capacitated_Capacity, MMNL_model_nonlinear;

''')

customer_segment = 2

M = list(range(1, customer_segment + 1))

product = 10

N = list(range(1, product + 1))

capacity = 1

K = [capacity]

gamma_ = np.divide(1, customer_segment)

gamma_= np.full(len(M),gamma_)

gamma = {

    (customer_segment_i): (gamma_[i])

    for i, customer_segment_i in enumerate(M)

}

rho_ = np.random.uniform(1, 3, (len(M), len(N)))

rho = {

    (customer_segment_i, product_j): (rho_[i][j])

    for i, customer_segment_i in enumerate(M)

    for j, product_j in enumerate(N)

}

v_ = np.random.uniform(0, 1, (len(M), len(N)))

v = {

    (customer_segment_i, product_j): (v_[i][j])

    for i, customer_segment_i in enumerate(M)

    for j, product_j in enumerate(N)

}

v0_ = [5, 5] 

v0 = {

    (customer_segment_i): (v0_[i])

    for i, customer_segment_i in enumerate(M)

}

Beta_ = np.random.uniform(0, 1, (len(K), len(N)))

Beta = {

    (capacity_k, product_j): (Beta_[k][j])

    for k, capacity_k in enumerate(K)

    for j, product_j in enumerate(N)

}

Capacity_Max = 5

ampl.set['M'] = M

ampl.set['N'] = N

ampl.set['K'] = K

ampl.param['gamma'] = gamma

ampl.param['rho'] = rho

ampl.param['v'] = v

ampl.param['v0'] = v0

ampl.param['Beta'] = Beta

ampl.eval('display M;')

ampl.eval('display N;')

ampl.eval('display K;')

ampl.eval('display gamma;')

ampl.eval('display rho;')

ampl.eval('display v;')

ampl.eval('display v0;')

ampl.eval('display Beta;')

Python code:

def f(CAOP_bilinear):

    dim = len(CAOP_bilinear)

    OF = 0

    x = 1

    for i in range(0, M):

        for j in range(0, N):

            t1 = 0

            t1 += gamma[i] * rho[i] * v[i] * y[i] * x[i]

            OF += t1

            return OF

    print(OF) 

[AMPL Google Group]

The value for the objective is zero before solving since all variables are zero at that point. You need to invoke ampl.solve() before you check the objective value:

ampl.solve()

ampl.eval("display Revenue;")

This will produce the following output:

Knitro 13.2.0: Locally optimal or satisfactory solution.
objective 0.9145105816; optimality gap 0
1 nodes; 1 subproblem solves

suffix feaserror OUT;
suffix opterror OUT;
suffix numfcevals OUT;
suffix numiters OUT;
suffix incumbent OUT;
suffix relaxbnd OUT;
Revenue = 0.914511

--
Filipe Brandão
am...@googlegroups.com

{#HS:2098669587-113260#}

[Phuc Nguyen]

Hey Filipe,

I managed to get it running. Thanks for your help!

Have a nice day.

--
You received this message because you are subscribed to a topic in the Google Groups "AMPL Modeling Language" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/ampl/U5w24rnwAEM/unsubscribe.
To unsubscribe from this group and all its topics, send an email to ampl+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ampl/reply-77152-2098669587-6244736235-1671188328-1521994610%40helpscout.net.

--

Best regards,

Phuc Nguyen

 

Tel: +84 932 673 939 (Whatsapp available)

Email: phucdn...@gmail.com