Cannot multiply a quadratic expression by an aff. expression

manoj.centura · May 26, 2022, 7:12pm

using  Cbc,JuMP,LinearAlgebra, Juniper,Ipopt,HiGHS;
optimizer = Juniper.Optimizer;
nl_solver = optimizer_with_attributes(Ipopt.Optimizer);

mip_solver = optimizer_with_attributes(HiGHS.Optimizer);
mod2 = Model(optimizer_with_attributes(optimizer, "nl_solver"=>nl_solver,
"mip_solver"=>mip_solver));
@variable(mod2,Tnk[h in 1:3,
                        i in 1:14])
@variable(mod2, 0.0 <= x[h in 1:5,
                        i in 1:3,
                        j in 1:2,
                        k in 1:1,
                        l in 1:2] <=19999);
mctc=[718.3492, 0.0237787, -6.845167e-7, 1.028082e-11, -6.545553000000001e-17, 1.472553e-22, 0.0, 0.0, 0.0];
trimhc=[196.3769 3.780828e-5 3.682656e-11 0.0 0.0 0.0 0.0 0.0; 161.5506 -4.104959e-5 1.164278e-10 0.0 0.0 0.0 0.0 0.0; 124.5838 -1.5003e-6 8.643112e-12 0.0 0.0 0.0 0.0 0.0; 87.61682 4.037087e-5 -1.147064e-10 0.0 0.0 0.0 0.0 0.0; 53.35888 -7.157266e-5 -2.043487e-10 0.0 0.0 0.0 0.0 0.0]
lcbc=[125.7708, -6.765074e-5, 1.182517e-9, -1.564492e-14, 5.572839e-20, 0.0, 0.0, 0.0];
zgrade=[1.1273957158962795 1.1547344110854503];
Tnkcn=[8.880501 -0.01480475 2.180284e-5 1.4783610000000008e-8 3.5880200000000023e-12 0.0 0.0 0.0;
 219.3388 0.004594099 3.936225e-6 1.4076890000000013e-9 1.7606220000000012e-13 0.0 0.0 0.0;
  195.0587 0.008311252 -2.186105e-5 2.339001e-9 -1.022953e-12 1.550898e-16 0.0 0.0;
  195.0587 0.008311252 2.186105e-5 2.3390010000000013e-8 1.0229530000000006e-11 1.5508980000000014e-15 0.0 0.0;
   161.5083 0.0001058785 4.817396000000002e-8 7.101192000000004e-12 0.0 0.0 0.0 0.0;
   161.5083 0.0001058785 4.817396000000002e-8 7.101192000000004e-12 0.0 0.0 0.0 0.0;
   124.582 -6.062317e-5 1.4651240000000006e-7 7.352221000000004e-11 1.0783470000000007e-14 0.0 0.0 0.0;
   124.582 -6.062317e-5 1.4651240000000006e-7 7.352221000000004e-11 1.0783470000000007e-14 0.0 0.0 0.0;
    87.87159 -0.0001834451 8.704129000000005e-8 1.755314000000001e-11 0.0 0.0 0.0 0.0; 87.87159 -0.0001834451 8.704129000000005e-8 1.755314000000001e-11 0.0 0.0 0.0 0.0; 53.56956 -0.001285654 1.11241e-6 5.177961000000003e-10 0.0 0.0 0.0 0.0; 53.56956 -0.001285654 1.11241e-6 5.177961000000003e-10 0.0 0.0 0.0 0.0; 44.81982 0.2890156 0.003176682 1.393608e-5 2.679035e-8 1.877435e-11 0.0 0.0; 44.81982 0.2890156 0.003176682 1.393608e-5 2.679035e-8 1.877435e-11 0.0 0.0]
ztrm=[0.97,0.97,0.97,0.97,0.97]
for p in 1:3
    DL=@expression(mod2,sum(x[h,i,j,k,l] for h in 1:5, i in 1:p, j in 1:2,k in 1:1, l in 1:2));
    mctcexp=@NLexpression(mod2,
           mctc[1] + mctc[2]*(DL*ztrm[p])
           + mctc[3]*(DL*ztrm[p])^2        + mctc[4]*(DL*ztrm[p])^3
           + mctc[5]*(DL*ztrm[p])^4        + mctc[6]*(DL*ztrm[p])^5
           + mctc[7]*(DL*ztrm[p])^6        + mctc[8]*(DL*ztrm[p])^7)
    lcbexp=@NLexpression(mod2, lcbc[1] + lcbc[2]*(DL*ztrm[p])
        + lcbc[3]*(DL*ztrm[p])^2
            + lcbc[4]*(DL*ztrm[p])^3        + lcbc[5]*(DL*ztrm[p])^4
            + lcbc[6]*(DL*ztrm[p])^5        + lcbc[7]*(DL*ztrm[p])^6
            + lcbc[8]*(DL*ztrm[p])^7
           )
Ntnknumber=10
zz=1/1.025
           WBMOM=@variable(mod2,[z in 1:Ntnknumber])
           for t in 1:Ntnknumber
           	   @NLconstraint(mod2,WBMOM[t]==Tnk[p,t]*(Tnkcn[t,1] + Tnkcn[t,2]*zz*Tnk[p,t]
           	   +  Tnkcn[t,3]*(zz*Tnk[p,t])^2 +Tnkcn[t,4]*(zz*Tnk[p,t])^3 +Tnkcn[t,5]*(zz*Tnk[p,t])^4
           	   +Tnkcn[t,6]*(zz*Tnk[p,t])^5 +Tnkcn[t,7]*(zz*Tnk[p,t])^6	))
           end
           ziter=0
           Ndimen=5*p*3*1*2
           @variable(mod2,HoldLMOM[z in 1:Ndimen])

           for h in 1:5
           	for i in 1:p
           		for j in 1:2
           		 for k=1:1
           			for l in 1: 2

           				ziter+=1
           				@constraint(mod2,HoldLMOM[ziter] == x[h,i,j,k,l]*
           				    		(trimhc[h,1]+trimhc[h,2]*zgrade[k,l]*x[h,i,j,k,l]
           				    		+trimhc[h,3]*(zgrade[k,l]*x[h,i,j,k,l])^2
           				   		+trimhc[h,4]*(zgrade[k,l]*x[h,i,j,k,l])^3
           				    		+trimhc[h,5]*(zgrade[k,l]*x[h,i,j,k,l])^4
           				    		+trimhc[h,6]*(zgrade[k,l]*x[h,i,j,k,l])^5
           				    		+trimhc[h,7]*(zgrade[k,l]*x[h,i,j,k,l])^6
           				    		+trimhc[h,8]*(zgrade[k,l]*x[h,i,j,k,l])^7))

           						end
           					end
           				end
           			end
           		end
           		println("LightshipMOM",LightshipMOM[p])
           		println("WCMOM",WCMOM[p])
           		println("HoldLMOM=",HoldLMOM[1])
           		println("HoldLMOM=",HoldLMOM[2])
           		println("HoldLMOM=",HoldLMOM[1])
           		println("HoldLMOM=",HoldLMOM[2])
           		@NLconstraint(mod2,LightshipMOM[p]+ sum(HoldLMOM[yy] for yy in 1:Ndimen) +
           		sum(WBMOM[yy] for yy in 1:Ntnknumber)+
           					   WCMOM[p] + WB3BHMOM ==
           		 		  -100*TRIM*mctcexp + lcbexp*DL)
end


oadError: Cannot multiply a quadratic expression by an aff. expression
in expression starting at D:\BCAP_Phase_2\BCAP_phase_4\bcap_Dynamic_expressions\dynamicexp.jl:29
error(s::String) at error.jl:33
*(q::QuadExpr, a::AffExpr) at operators.jl:305
promote_operation_fallback(op::typeof(*), #unused#::Type{QuadExpr}, #unused#::Type{AffExpr}) at interface.jl:37
promote_operation(::typeof(*), ::Type, ::Type) at interface.jl:99
promote_operation_fallback(::typeof(*), ::Type{AffExpr}, ::Type{AffExpr}, ::Type{AffExpr}) at interface.jl:56
promote_operation(::typeof(*), ::Type, ::Type, ::Type) at interface.jl:99
promote_operation_fallback(#unused#::typeof(*), #unused#::Type{VariableRef}, #unused#::Type{Float64}, #unused#::Type{AffExpr}, args::Type) at interface.jl:56
promote_operation(::typeof(*), ::Type, ::Type, ::Type, ::Type) at interface.jl:99
promote_operation_fallback(::typeof(MutableArithmetics.sub_mul), ::Type, ::Type, ::Type, ::Type, ::Type) at interface.jl:85
promote_operation(::typeof(MutableArithmetics.sub_mul), ::Type, ::Type, ::Type, ::Type, ::Type) at interface.jl:99
mutability(::Type, ::Function, ::Type, ::Type, ::Type, ::Type, ::Type) at interface.jl:232
mutability at interface.jl:240 [inlined]
operate!!(::typeof(MutableArithmetics.sub_mul), ::QuadExpr, ::VariableRef, ::Float64, ::AffExpr, ::AffExpr) at rewrite.jl:88
macro expansion at rewrite.jl:289 [inlined]
macro expansion at macros.jl:815 [inlined]
top-level scope at dynamicexp.jl:61
eval at boot.jl:360 [inlined]
include_string(mapexpr::typeof(identity), mod::Module, code::String, filename::String) at loading.jl:1116

odow · May 26, 2022, 7:50pm

Yes, you cannot multiply a quadratic expression by an affine expression in the normal macros. Use @NL instead:

julia> model = Model();

julia> @variable(model, x)
x

julia> quad = @expression(model, x^2)
x²

julia> aff = @expression(model, x + 1)
x + 1

julia> @expression(model, quad * aff)
ERROR: Cannot multiply a quadratic expression by an aff. expression
Stacktrace:
 [1] error(s::String)
   @ Base ./error.jl:33
 [2] *(q::QuadExpr, a::AffExpr)
   @ JuMP ~/.julia/packages/JuMP/Psd1J/src/operators.jl:305
 [3] operate(::typeof(*), ::QuadExpr, ::AffExpr)
   @ MutableArithmetics ~/.julia/packages/MutableArithmetics/L8iac/src/interface.jl:186
 [4] operate(::typeof(MutableArithmetics.add_mul), ::MutableArithmetics.Zero, ::QuadExpr, ::AffExpr)
   @ MutableArithmetics ~/.julia/packages/MutableArithmetics/L8iac/src/rewrite.jl:40
 [5] operate_fallback!!(::MutableArithmetics.IsNotMutable, ::typeof(MutableArithmetics.add_mul), ::MutableArithmetics.Zero, ::QuadExpr, ::AffExpr)
   @ MutableArithmetics ~/.julia/packages/MutableArithmetics/L8iac/src/interface.jl:555
 [6] operate!!(::typeof(MutableArithmetics.add_mul), ::MutableArithmetics.Zero, ::QuadExpr, ::AffExpr)
   @ MutableArithmetics ~/.julia/packages/MutableArithmetics/L8iac/src/rewrite.jl:94
 [7] macro expansion
   @ ~/.julia/packages/MutableArithmetics/L8iac/src/rewrite.jl:294 [inlined]
 [8] macro expansion
   @ ~/.julia/packages/JuMP/Psd1J/src/macros.jl:1352 [inlined]
 [9] top-level scope
   @ REPL[87]:1

julia> @NLexpression(model, quad * aff)
subexpression[1]: (0 + x * x) * (1.0 + x)

manoj.centura · May 27, 2022, 10:34am

oh i missed using @NL macro

Topic		Replies	Views
Alpine+JuMP gives error for quadratic functions Optimization (Mathematical)	4	990	September 7, 2022
Cannot multiply a quadratic expression by an affine expression Optimization (Mathematical)	6	500	July 7, 2023
Multiplication of matrix expression and variables leads to stack overflow and matmul error Optimization (Mathematical) jump , optimization , stackoverflow , nonlinear	2	232	March 18, 2024
Juniper integer-objective fails on specific parameters Optimization (Mathematical)	6	328	March 16, 2022
JuMP Non-Linear Optimization Optimization (Mathematical)	14	696	February 22, 2023

Cannot multiply a quadratic expression by an aff. expression

Related topics