[R] gss package - predict.ssanova

[José Ernesto Jardim]

Chong Gu wrote:

>    Sender: ernesto at neptuno.ipimar.pt
>    Date: Tue, 08 May 2001 09:51:08 +0100
>    From: =?iso-8859-1?Q?Jos=E9?= Ernesto Jardim <ernesto at ipimar.pt>
>    Organization: IPIMAR
>    X-Mailer: Mozilla 4.72 [en] (X11; I; Linux 2.2.14 i686)
>    X-Accept-Language: en
>    MIME-Version: 1.0
>    CC: R-help at stat.math.ethz.ch
>    References: <200105071639.LAA88280 at odds.stat.purdue.edu>
>    Content-Type: text/plain; charset=us-ascii
>    Content-Transfer-Encoding: 7bit
>
>    Chong Gu wrote:
>
>    >    Here is the information about this objects.
>    >
>    >    > recs.spliney
>    >
>    >    Call:
>    >    ssanova(formula = recs.mds[[1]][, 2] ~ recs.loc$x * recs.loc$y,     type = "tp")
>    >
>    >    Terms:
>    >    [1] "1"                     "recs.loc$x"            "recs.loc$y"
>    >    [4] "recs.loc$x:recs.loc$y"
>    >
>    >    Number of fixed and random effects:
>    >
>    >                          Fixed Random
>    >    1                         1      0
>    >    recs.loc$x                1      1
>    >    recs.loc$y                1      1
>    >    recs.loc$x:recs.loc$y     1      3
>    >    total                     4      5
>    >
>    >    Smoothing parameters are selected by GCV.
>    >
>    >    > is.data.frame(recn.grid)
>    >    [1] TRUE
>    >    > names(recn.grid)
>    >    [1] "x" "y"
>    >
>    >    Regards
>    >
>    >    EJ
>    >
>    > The variables in the fit are "recs.loc$x" and "recs.loc$y", but the
>    > variables in recn.grid are "x" and "y".
>    >
>    > BTW, are recs.loc$x and recs.loc$y the x-y coordinate of geographic
>    > locations?  If the answer is yes, then you probably want to enter it
>    > as a single variable with two columns.  What you are now doing appears
>    > to fit a tensor product spline.
>    >
>    > -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
>    > r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
>    > Send "info", "help", or "[un]subscribe"
>    > (in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
>    > _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
>
>    Hi
>
>    I've tried your proposal and I got an error message
>
>    > ssanova(recn.mds[[1]][,1]~recn.loc,"tp")
>    Error in model.frame(formula, rownames, variables, varnames, extras, extranames,  :
>            invalid variable type
>
>    The objects are in aggreement with your description
>
>    > is.data.frame(recn.loc)
>    [1] TRUE
>    > is.vector(recn.mds[[1]][,1])
>    [1] TRUE
>    > names(recn.loc)
>    [1] "x" "y"
>
>    Regards
>
>    EJ
>
> recn.loc should be a matrix, not a data frame.

Hi

I've set recn.loc as a matrix and it worked but I'm still getting a error message from
predict.ssanova

> predict.ssanova(recs.splinex,recn.grid,se.fit=FALSE)
Error: subscript out of bounds
> colnames(recn.loc)
[1] "x" "y"
> colnames(recn.grid)
[1] "x" "y"
> is.matrix(recn.loc)
[1] TRUE
> is.matrix(recn.grid)
[1] TRUE

If I keep recn.grid has a data.frame I get

> predict.ssanova(recs.splinex,recn.grid,se.fit=FALSE)
Error in array(x, c(length(x), 1), if (!is.null(names(x))) list(names(x),  :
        attempt to set an attribute on NULL


Regards

EJ

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._