[R] model fine, predict gives an error

[Antje Niederlein]

Yes, I did send an attachment - but I forgot that attachments might be
removed, no?

Here it is:

http://rapidshare.com/files/452815636/drmData.RData

Antje




On 16 March 2011 12:19, Mike Marchywka <marchywka at hotmail.com> wrote:
>
>
>
>
>
>
>
> ----------------------------------------
>> Date: Wed, 16 Mar 2011 08:17:49 +0100
>> From: niederlein-rstat at yahoo.de
>> To: r-help at r-project.org
>> Subject: Re: [R] model fine, predict gives an error
>>
>> Anybody who can help me with this issue?
>
> did you post your data? I was curious so I thought I may look
> but you need to post data so people can determine if they
> can produce and fix your problem. Any complaints
> about bad/ singular matricies are likely to depend on data
> being bad in some way.
>
>
>
>
>>
>>
>>
>>
>> On 15 March 2011 14:15, Antje Niederlein  wrote:
>> > Hi there,
>> >
>> > I try to model some dose response curves (drc-package). In most cases
>> > it is fine but now I got some data which produces me the following
>> > error:
>> >
>> > load("drmData.RData")
>> > library(drc)
>> >
>> > drmObj <- drm(value ~ concentration, cmpd_respvar, data = drmData, fct = LL.4())
>> > predict(drmObj)
>> >
>> >>> Error in chol.default(0.99 * object$fit$hessian + 0.01 * diag(dim(object$fit$hessian)[1])) :
>> >  the leading minor of order 2 is not positive definite
>> > Error in resultMat[, 2] <- estSE : replacement has length zero
>> >
>> > Can anybody explain to me why the calculation of the dose response
>> > model works but the prediction fails? Anything I can do in this case?
>> >
>> > Antje
>> >
>>
>> ______________________________________________
>> 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.
>