[BioC] express & summary.stat=li.wong -> error report
cberry@tajo.ucsd.edu
cberry@tajo.ucsd.edu
Thu, 14 Nov 2002 08:50:00 -0800 (PST)
It does.
But the patch I suggested got Hinnerk past the first one (1127 in
sequence).
4318 appears to be "U34301_at" which has 2 probe sets.
On Thu, 14 Nov 2002, Rafael A. Irizarry wrote:
> i think the hg6800 chip has a one probe probe set which makes li.wong
> break down.
>
> On Thu, 14 Nov 2002, Hinnerk Boriss wrote:
>
> > Yes, I do work with Hu6800 chips and thanks, I get past that error
> > message, however, I run into a new one:
> > Error in apply(resid[, !phi.outliers]^2, 1, sum) :
> > dim(x) must have a positive length
> >
> > > traceback()
> > 8: stop("dim(X) must have a positive length")
> > 7: apply(resid[, !phi.outliers]^2, 1, sum)
> > 6: fit.li.wong(t(data.matrix), remove.outliers, normal.array.quantile,
> > normal.resid.quantile, large.threshold, large.variation,
> > outlier.fraction, delta, maxit, outer.maxit, verbose)
> > 5: FUN(X[[4318]], ...)
> > 4: lapply(as.list(X), FUN, ...)
> > 3: sapply(data.lst, summary.stat, ...)
> > 2: t(sapply(data.lst, summary.stat, ...))
> > 1: express(NormLemonData, summary.stat = li.wong)
> >
> >
> > Any suggestions on that one?
> >
> > Cheers,
> > Hinnerk
> >
> >
> > >
> > > I think that error message is what I got in using an Hu6800 chip where
> > > there are (apparently) two probe sets with just one pair. The
> > > hack I used
> > > was to insert after this line in fit.li.wong:
> > >
> > > J <- dim(data.matrix)[2]
> > >
> > > the following comment and line:
> > >
> > > ### The following line is intended to deal with a problem in the
> > > ### treatment of Hu6800 chips. That chip has a couple of probe sets
> > > ### in which there is only one PM-MM pair. Probably, it is just a
> > > ### control region, so I am omitting the fit.li.wong operations for
> > > ### any such probe set
> > >
> > > if (J == 1)
> > > return(list(theta=c(data.matrix),sigma.theta=rep(NA,length(dat
> > a.matrix))))
> > >
> > >
> > >
> > > On Wed, 13 Nov 2002, Rafael A. Irizarry wrote:
> > >
> > > > my guess is that one of those guys was left with just one
> > > column (or
> > > > none) and was turned into a vector. we probably need to add
> > > a drop=FALSE
> > > > somewhere.
> > > >
> > > > On Wed, 13 Nov 2002, Hinnerk Boriss wrote:
> > > >
> > > > > Hi there!
> > > > >
> > > > > I have a statement like this:
> > > > >
> > > > > data.lwf <- express(data, summary.stat=li.wong)
> > > > >
> > > > > and I get the following error message:
> > > > >
> > > > > Error in data.matrix[,!phi.outliers] %*% phi[!phi.outliers] :
> > > > > non-conformable arguments
> > > > >
> > > > > What does this error message mean in detail? Does the
> > > li.wong routine
> > > > > require to do some outlier checking beforehand? If so,
> > > what method is
> > > > > approporiate? Just to mention, I am not a fan of removing
> > > outliers.
> > > > > Anyway, How would I find the trouble making values in my data?
> > > > >
> > > > > Cheers,
> > > > > Hinnerk
> > > > >
> > > > >
> > > > > _______________________________________________
> > > > > Bioconductor mailing list
> > > > > Bioconductor@stat.math.ethz.ch
> > > > > http://www.stat.math.ethz.ch/mailman/listinfo/bioconductor
> > > > >
> > > >
> > > > _______________________________________________
> > > > Bioconductor mailing list
> > > > Bioconductor@stat.math.ethz.ch
> > > > http://www.stat.math.ethz.ch/mailman/listinfo/bioconductor
> > > >
> > >
> > > Charles C. Berry (858) 534-2098
> > > Dept of
> > > Family/Preventive Medicine
> > > E mailto:cberry@tajo.ucsd.edu UC San Diego
> > > http://hacuna.ucsd.edu/members/ccb.html La Jolla, San Diego
> > > 92093-0645
> > >
> > >
> > >
> > > _______________________________________________
> > > Bioconductor mailing list
> > > Bioconductor@stat.math.ethz.ch
> > > http://www.stat.math.ethz.ch/mailman/listinfo/bioconductor
> > >
> >
> >
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor@stat.math.ethz.ch
> http://www.stat.math.ethz.ch/mailman/listinfo/bioconductor
>
Charles C. Berry (858) 534-2098
Dept of Family/Preventive Medicine
E mailto:cberry@tajo.ucsd.edu UC San Diego
http://hacuna.ucsd.edu/members/ccb.html La Jolla, San Diego 92093-0645