[R] biglm() and NeweyWest()

[Thomas Lumley]

On Tue, Jul 26, 2011 at 12:29 AM, Philipp Grueber
<philipp.grueber at ebs.edu> wrote:
> Dear all,
>
> I am working on a large dataset and need to use biglm() to perform OLS
> regressions. I have detected significant ARCH effects which I try to account
> for using the Newey-West correction.
>
> So far, I have worked with NeweyWest() in the sandwich package. NeweyWest()
> however seems to be unable to handle an object of class "biglm".
>

An object of class 'biglm' doesn't contain enough information to
compute the Newey-West estimator. The whole point of biglm() is that
it works with one observation at a time and doesn't store all the
data.  You'll probably need to do nearly all the work for the
Newey-West estimator yourself -- you can get the fitted values and the
information matrix out of biglm(), but not much else.

  -thomas

-- 
Thomas Lumley
Professor of Biostatistics
University of Auckland