[Date Index] [Thread Index] [Author Index]

A good lag value for the Portmanteau test

• To: mathgroup at smc.vnet.net
• Subject: [mg66446] A good lag value for the Portmanteau test
• From: "john.hawkin at gmail.com" <john.hawkin at gmail.com>
• Date: Sat, 13 May 2006 02:42:33 -0400 (EDT)
• Sender: owner-wri-mathgroup at wolfram.com

Hello everyone,

I was wondering if anyone knows what the criteria is for choosing a
good lag value for the Portmanteau test (done on the residuals of a
fitted model).  The value of h=35 is suggested by Mathematica's
documentation for time series of around n=100 or n=150, but for time
series of length n=250 or n=1000 should this stay the same, or should
it be some percentage of n?

Thanks,

-John Hawkin