[R] Exception in NeweyWest - Pre-Whitening necessary?

[Achim Zeileis]

On Fri, 9 Sep 2011, Simon Zehnder wrote:

> Hi guyz,
>
> I have run my algorithm in R (see http://pastebin.com/q84Tujfg) and got 
> the following error:

This is not reproducible. There is only the script but no information 
about how it is called. Also, within your script you call 
NeweyWest(regression) -- however, "regression" is not defined within the 
script!

> Error in ar.ols(x, aic = aic, order.max = order.max, na.action = na.action,  :
>  'order.max' must be < 'n.used'

NeweyWest call ar() essentially with order.max = as.integer(prewhite) 
which defaults to 1. Thus, the error must be caused by a model with
n.used <= 1 which suggests that it is not a very sensible model to start 
with.

> I am pretty sure, that the error comes from the NeweyWest function in 
> line 45, as the NeweyWest function uses the ar.ols() function for pre 
> whitening. Does anyone has an idea how to circumvent this? Can I just 
> shut off pre whitening?

Why don't you read the documentation in ?NeweyWest or in 
vignette("sandwich", package = "sandwich")?
Z

> I am thankful for every suggestion.
>
> Simon
>
>
> 	[[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>