<hansl
mx =zeros(1,1000)|cum(mnormal(49,1000))
my =zeros(1,1000)|cum(mnormal(49,1000)) cnst = ones(50,1)
matrix Ubig = {}
matrix Vbig = {}
matrix bbig = {}
matrix U = {}
matrix V = {}
matrix b = {}
loop i=1..1000 -q
b =mols(my[,i],cnst~mx[,i],&U,&V)
bbig = bbig|b'
Ubig = Ubig|U'
Vbig = Vbig|V
endloop
hasl>
bbig ii store coeffs
Ubig - residuals
Vbig - vcovs
also type
help zeros
help cum
help mnormal etc
Oleh
14 грудня 2015, 01:18:54, від "Paolo Pocaterra" <poka92.ra(a)hotmail.it>:
Hello to everybody,
I’m Paolo Pocaterra and I’m a new user of this list. I’m in trouble with the Monte Carlo
simulation in gretl and I hope someone of you can help me. I need someone who is able to
clarify and explain the aim and mostly the code on grelt of the Monte Carlo simulations.
Actually, I’m trying to generate 1000 Monte Carlo simulations using a sample size of 50
(T=50).
Considering the following model
yt = b1yt−1 + ut, ut iid ∼ N(0,1)
xt = b2xt−1 + et, et iid ∼ N(0,1)
where t = 0,…,T; b1=1 and b2=1, y0 = 0 e x0 = 0.
and I have to estimate
yt = alpha+ xt + vt
where vt is the usual error term.
Thanks in advance,
Paolo Pocaterra.
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users