Dear all,
I need to minimize the following function L=(Xit*b- Y)^2 with nonlinear constraints.
I have imported the matrix Xit [556,28] and the vector Y[556]
so I need to find (the vector) b[28] subject to: b>=betaMAX (real colvector)
I wrote the following commands in mata but Stata reports a "redundant or inconsistent constraints".
This is my code:
mata
void maxme(todo,b,y,X,lnf,g,H)
{
lnf = (-y :+ X*b'):^2
}
init=J(cols(X),1,1)
s=optimize_init()
//build the constraint matrix
C = I(28)
c = exp(init):+betaMAX
Cc = (C, c)
optimize_init_constraints(s,Cc)
optimize_init_evaluator(s,&maxme())
optimize_init_evaluatortype(s,"v0")
optimize_init_argument(s, 1, y)
optimize_init_argument(s, 2, X)
optimize_init_evaluatortype(s,"d0")
optimize_init_which(s,"min")
optimize_init_params(s, init')
b = optimize(s)
b'
end
Thank you,
Giulia
I need to minimize the following function L=(Xit*b- Y)^2 with nonlinear constraints.
I have imported the matrix Xit [556,28] and the vector Y[556]
so I need to find (the vector) b[28] subject to: b>=betaMAX (real colvector)
I wrote the following commands in mata but Stata reports a "redundant or inconsistent constraints".
This is my code:
mata
void maxme(todo,b,y,X,lnf,g,H)
{
lnf = (-y :+ X*b'):^2
}
init=J(cols(X),1,1)
s=optimize_init()
//build the constraint matrix
C = I(28)
c = exp(init):+betaMAX
Cc = (C, c)
optimize_init_constraints(s,Cc)
optimize_init_evaluator(s,&maxme())
optimize_init_evaluatortype(s,"v0")
optimize_init_argument(s, 1, y)
optimize_init_argument(s, 2, X)
optimize_init_evaluatortype(s,"d0")
optimize_init_which(s,"min")
optimize_init_params(s, init')
b = optimize(s)
b'
end
Thank you,
Giulia