James Anderson wrote:
> I am trying to use pamr (prediction analysis of microarray), which
uses nearest shrunken centroid classification. Now I want to load my
data using read.table, then use pamr. the codes are as follows:
>
> Training = read.table("A.txt",header = FALSE)
> myData = list(x = Training,y = factor(TrainingLabel));
> mytrain = pamr.train(myData)
> It always complains "Error in x %*% Y : requires numeric
matrix/vector arguments"
>
> I think the main reason is variable Training read from read.table is
an object, not just numeric matrix, I tried as.numeric(Training), it
still has errors. I know if Training is created by something like
>
> Training = matrix(rnorm(1000*20),ncol=20)
> It will not complain, since by this way, Training is purely a
numeric matrix, not object. Does anybody know how to convert the
object read from read.table to a pure numeric matrix?
>
Hi, James.
If A.txt has any text columns in it, you will want to remove those
before reading it into R (or after it is in R). Then, you should be
able to do:
Training.matrix = as.matrix(Training)
Does that do what you want?
Sean
Quoting James Anderson <janderson_net at="" yahoo.com="">:
> I am trying to use pamr (prediction analysis of microarray), which
> uses nearest shrunken centroid classification. Now I want to load my
> data using read.table, then use pamr. the codes are as follows:
>
> Training = read.table("A.txt",header = FALSE)
> myData = list(x = Training,y = factor(TrainingLabel));
> mytrain = pamr.train(myData)
> It always complains "Error in x %*% Y : requires numeric
> matrix/vector arguments"
>
> I think the main reason is variable Training read from read.table is
> an object, not just numeric matrix, I tried as.numeric(Training),
> it still has errors. I know if Training is created by something
like
>
> Training = matrix(rnorm(1000*20),ncol=20)
> It will not complain, since by this way, Training is purely a
> numeric matrix, not object. Does anybody know how to convert the
> object read from read.table to a pure numeric matrix?
>
> Thanks,
>
> James
I am not familiar with pamr and the arguments it requires, but
read.table will create a data frame object. As it has no header and
(presumably) it only contains numbers in a table form, if you use
'as.matrix(Training)' you'll convert Training from a data frame to a
numeric matrix. If pamr expects a list where the first element is a
matrix, it may work. Anyway, use 'as.matrix', not 'as.numeric'
I hope this helps a bit!
Jose
--
Dr. Jose I. de las Heras Email: J.delasHeras at
ed.ac.uk
The Wellcome Trust Centre for Cell Biology Phone: +44 (0)131
6513374
Institute for Cell & Molecular Biology Fax: +44 (0)131
6507360
Swann Building, Mayfield Road
University of Edinburgh
Edinburgh EH9 3JR
UK
If you look at the documentation for read.table, you will see a
variable " colClasses". You can use this to set each column to the
correct type.
--Naomi
At 09:34 PM 2/24/2007, J.delasHeras at ed.ac.uk wrote:
>Quoting James Anderson <janderson_net at="" yahoo.com="">:
>
> > I am trying to use pamr (prediction analysis of microarray), which
> > uses nearest shrunken centroid classification. Now I want to load
my
> > data using read.table, then use pamr. the codes are as follows:
> >
> > Training = read.table("A.txt",header = FALSE)
> > myData = list(x = Training,y = factor(TrainingLabel));
> > mytrain = pamr.train(myData)
> > It always complains "Error in x %*% Y : requires numeric
> > matrix/vector arguments"
> >
> > I think the main reason is variable Training read from read.table
is
> > an object, not just numeric matrix, I tried as.numeric(Training),
> > it still has errors. I know if Training is created by something
like
> >
> > Training = matrix(rnorm(1000*20),ncol=20)
> > It will not complain, since by this way, Training is purely a
> > numeric matrix, not object. Does anybody know how to convert the
> > object read from read.table to a pure numeric matrix?
> >
> > Thanks,
> >
> > James
>
>I am not familiar with pamr and the arguments it requires, but
>read.table will create a data frame object. As it has no header and
>(presumably) it only contains numbers in a table form, if you use
>'as.matrix(Training)' you'll convert Training from a data frame to a
>numeric matrix. If pamr expects a list where the first element is a
>matrix, it may work. Anyway, use 'as.matrix', not 'as.numeric'
>
>I hope this helps a bit!
>
>Jose
>
>--
>Dr. Jose I. de las Heras Email: J.delasHeras at
ed.ac.uk
>The Wellcome Trust Centre for Cell Biology Phone: +44 (0)131
6513374
>Institute for Cell & Molecular Biology Fax: +44 (0)131
6507360
>Swann Building, Mayfield Road
>University of Edinburgh
>Edinburgh EH9 3JR
>UK
>
>_______________________________________________
>Bioconductor mailing list
>Bioconductor at stat.math.ethz.ch
>https://stat.ethz.ch/mailman/listinfo/bioconductor
>Search the archives:
>http://news.gmane.org/gmane.science.biology.informatics.conductor
Naomi S. Altman 814-865-3791 (voice)
Associate Professor
Dept. of Statistics 814-863-7114 (fax)
Penn State University 814-865-1348
(Statistics)
University Park, PA 16802-2111
