Hi everyone, I’ve modeled a simple model to assign instruments to procedures. I’ve implemented the model in JuMP using the solver Gurobi and package MOA.jl.
When I tried to solve the model, the message below appear to me:
Gurobi Optimizer version 10.0.0 build v10.0.0rc2 (win64)
CPU model: Intel(R) Core(TM) i5-8265U CPU @ 1.60GHz, instruction set [SSE2|AVX|AVX2]
Thread count: 4 physical cores, 8 logical processors, using up to 8 threads
Optimize a model with 94 rows, 128 columns and 678 nonzeros
Model fingerprint: 0x2ada338b
Variable types: 10 continuous, 118 integer (6 binary)
Coefficient statistics:
Matrix range [5e-02, 9e+01]
Objective range [1e-01, 5e-01]
Bounds range [0e+00, 0e+00]
RHS range [1e+00, 6e+00]
Found heuristic solution: objective 8.7745536
Presolve removed 34 rows and 34 columns
Presolve time: 0.00s
Presolved: 60 rows, 94 columns, 430 nonzeros
Variable types: 0 continuous, 94 integer (9 binary)
Found heuristic solution: objective 8.2703373
Root relaxation: objective 7.760417e+00, 47 iterations, 0.00 seconds (0.00 work units)
Nodes | Current Node | Objective Bounds | Work
Expl Unexpl | Obj Depth IntInf | Incumbent BestBd Gap | It/Node Time
* 0 0 0 7.7604167 7.76042 0.00% - 0s
Explored 1 nodes (47 simplex iterations) in 0.01 seconds (0.00 work units)
Thread count was 8 (of 8 available processors)
Solution count 3: 7.76042 8.27034 8.77455
Optimal solution found (tolerance 1.00e-04)
Best objective 7.760416666667e+00, best bound 7.760416666667e+00, gap 0.0000%
User-callback calls 367, time in user-callback 0.00 sec
ERROR: In `MathOptInterface.ScalarAffineFunction{Float64}`-in-`MathOptInterface.LessThan{Float64}` constraint: Constant -1.25 of the function is not zero. The function constant should be moved to the set. You can use `MOI.Utilities.normalize_and_add_constraint` which does this automatically.
I need help to fix this error and solve my problem.
The model was defined as follows:
using JuMP, Gurobi
import DelimitedFiles
import MultiObjectiveAlgorithms as MOA
#Data reading
data = DelimitedFiles.readdlm(filename.txt)
instruments = data[1,1]
trays = 6
surgeries = data[1,3]
capacity = 89 #data[1,5]
freq = zeros(Float64, surgeries)
for i = 1:surgeries
freq[i] = data[2,i+1]
end
demand_max = zeros(Int, instruments)
for i = 1:instruments
demand_max[i] = data[3,i+1]
end
demand = zeros(Int, instruments, surgeries)
for i = 1:instruments
for j = 1:surgeries
demand[i,j] = data[i+3,j]
end
end
#Normalizing the frequency
max_freq = 0
min_freq = 9999999
for i = 1:surgeries
if(freq[i] > max_freq)
max_freq = freq[i]
end
if(freq[i] < min_freq)
min_freq = freq[i]
end
end
for i = 1:surgeries
freq[i] = (freq[i] - min_freq)/(max_freq - min_freq)
end
#Creating the model and declaring variables
#model = Model()
#model = Model(Gurobi.Optimizer)
model = Model(() -> MOA.Optimizer(Gurobi.Optimizer))
set_attribute(model, MOA.Algorithm(), MOA.EpsilonConstraint())
#set_optimizer_attribute(model, "TimeLimit", 600)
#set_optimizer_attribute(model, "Presolve", 0)
@variable(model, y[k = 1:trays], Bin) #(Equação 9)
@variable(model, x[i = 1:instruments, k = 1:trays] >= 0, Int) #(Equação 10)
@variable(model, e[i = 1:instruments, j = 1:surgeries] >= 0, Int) #(Equação 11.1)
@variable(model, n[i = 1:instruments, j = 1:surgeries] >= 0, Int) #(Equação 11.2)
@variable(model, a[j = 1:surgeries] >= 0) #(Equação 12.1)
@variable(model, b[j = 1:surgeries] >= 0) #(Equação 12.2)
#Defining the constraints
for i = 1:instruments, j = 1:surgeries
@constraint(model, e[i,j] >= demand[i,j] - sum(x[i,k] for k = 1:trays)) #(Equação 3)
@constraint(model, n[i,j] >= sum(x[i,k] for k = 1:trays) - demand[i,j]) #(Equação 4)
end
for j = 1:surgeries
#(Equação 5)
@constraint(model, a[j] == sum((demand[i,j] - e[i,j]) for i = 1:instruments)/sum(demand[i,j] for i = 1:instruments))
#(Equação 6)
@constraint(model, b[j] == sum(n[i,j] for i = 1:instruments)/sum((demand_max[i] - demand[i,j]) for i = 1:instruments))
end
for k = 1:trays
@constraint(model, sum(x[i,k] for i = 1:instruments) <= capacity*y[k]) #(Equação 7.1)
@constraint(model, sum(x[i,k] for i = 1:instruments) >= 0.5*capacity*y[k]) #(Equação 7.2)
end
@constraint(model, sum(y[k] for k = 1:trays) <= 6) #(Equação 8.1)
@constraint(model, sum(y[k] for k = 1:trays) >= 3) #(Equação 8.2)
#Objective Functions
@expression(model, instruments_expr, sum(x[i,k] for i = 1:instruments, k = 1:trays))
@expression(model, eff_expr, sum((freq[j]*((1 - b[j] + a[j])/2)) for j = 1:surgeries))
@expression(model, a_mean_expr, sum(a[j] for j=1:surgeries)/surgeries)
@expression(model, b_mean_expr, sum(b[j] for j=1:surgeries)/surgeries)
@expression(model, nb_trays_expr, sum(y[k] for k=1:trays))
@objective(model, Min, [-eff_expr, instruments_expr])
#Running the solver
optimize!(model)
status = termination_status(model)
println("STATUS: ", status, " ---------------------------")
print(solution_summary(model))
A simple instance of the problem is defined as follows:
7 Instruments, 5 Procedures, 15 Capacity
Procedures_Freq 10 11 12 13 14
Instruments_DemMax 4 4 4 5 4 5 4
2 4 3 1 1
4 2 3 0 2
3 4 0 5 2
1 2 5 3 4
4 1 3 4 1
5 3 0 1 2
0 2 4 2 0
. I should have caught this earlier, sorry for that.)