From ignacio.diaz-emparanza@ehu.es Thu May 27 12:22:15 2010
From: Ignacio Diaz-Emparanza <ignacio.diaz-emparanza@ehu.es>
To: gretl-devel@gretlml.univpm.it
Subject: [Gretl-devel] BIC in ols
Date: Thu, 27 May 2010 18:22:07 +0200
Message-ID: <1274977327.2116.15.camel@U001082>
In-Reply-To: AANLkTilV4-H-y0EzTzPgNIWcpASMw7SOoHoTJxEgF76b@mail.gmail.com
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="===============1817177396273263501=="

--===============1817177396273263501==
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: 8bit

I am a little confused about the behaviour of AIC, HQC and BIC in 
the standard ols command. I estimated some AR(p) models with different p
values with the 'ols' and the 'var' commands. The last with only one
variable which I know it is not designed for, but ... it works. 

I see that the BIC obtained through both commands:

1- are not the same. I suspect may be some different scale factor in the
formulas but this should not be a problem.

2- leads to different specification, and yes, this is a problem. In
concrete, the BIC in ols would suggest the maximum number of lags, and
this does not seem to be correct if one have a look at the correlograms.

<script>
#AR(p) selection 
function void ARF(series y, scalar maxorder)
    matrix C=zeros(maxorder,3) 
    loop i=1..maxorder
        var i y
        genr C[i,1]=i
        genr C[i,2]=$bic
        list xlist = lags(i,y)
        ols y 0 xlist
        genr C[i,3]=$bic
    end loop
C
end function

open data9-3.gdt
Y=diff(reskwh)
Y=sdiff(Y)
ARF(Y, 30)
</script>

[Linux Ubuntu 10.04 and gretl from CVS build date 2010-05-14]


-- 
Ignacio Diaz-Emparanza  
DEPARTAMENTO DE ECONOMÍA APLICADA III (ECONOMETRÍA Y ESTADÍSTICA)
UPV/EHU
Avda. Lehendakari Aguirre, 83 | 48015 BILBAO
T.: +34 946013732 | F.: +34 946013754
www.ea3.ehu.es 



--===============1817177396273263501==--