using optimize() to solve optimal weights

[Sarah Yang]

Hello,

i'm trying to solve the following optimization problem:

max F=x7*(v1*x1+v2*x2+v3*x3+v4*x4+v5*x5+v6*x6)^-1

x1-x7: financial variables

v1-v6: weights which i 'm trying to solve to make the "F" to maximize

F: it takes a value between 0 and 1

here is the code which i just revise the code from stata's example

Code:

mata:
A=st_data(.,"x1")
B=st_data(.,"x2")
C=st_data(.,"x3")
D=st_data(.,"x4")
E=st_data(.,"x5")
I=st_data(.,"x6")
J=st_data(.,"x7")

void solve(todo,v, f, g, H)
{
external A,B,C,D,E,I,J
f = J*(v[1]'*A +v[2]'*B +v[3]'*C +v[4]'*D +v[5]'*E +v[6]'*I)^-1
}

S=optimize_init()
optimize_init_evaluator(S,&solve())
optimize_init_params(S,(0))
optimize_init_which(S,"max")
v=optimize(S)
v
end


stata shows that there is an error: <istmt>: 3499 S not found

Could someone tell me where is the problem? or how do i revise the code?

thank you in advance

[Matthew J. Baker]

Sarah Yang --

Your code on the surface seems fine (barring some problem with identification). The one thing is you might want to fix the negative exponent so you have:

f = J*(v[1]'*A +v[2]'*B +v[3]'*C +v[4]'*D +v[5]'*E +v[6]'*I)^(-1)

Also, sometimes I find that Stata runs "too fast" through the model statement, so breaking it up so that you do:

mata:
A=st_data(.,"x1")
B=st_data(.,"x2")
C=st_data(.,"x3")
D=st_data(.,"x4")
E=st_data(.,"x5")
I=st_data(.,"x6")
J=st_data(.,"x7")

void solve(todo,v, f, g, H)
{
external A,B,C,D,E,I,J
f = J*(v[1]'*A +v[2]'*B +v[3]'*C +v[4]'*D +v[5]'*E +v[6]'*I)^-1
}
end

mata:
S=optimize_init()
optimize_init_evaluator(S,&solve())
optimize_init_params(S,(0))
optimize_init_which(S,"max")
v=optimize(S)
v
end

This sometimes helps, oddly.

Matthew J. Baker