Hej Fereshteh Izadi!
At the moment, your mycounts object is a data.frame mixing both characters (the first column) and integer/numeric values (your second to fourth columns).
Depending on what you want to do with your object, e.g. coerce it as a matrix using as.matrix(mycounts) you will get the error you see, because of the different datatypes in your data.frame.
As you have not included all the commands you ran, including the one generating the error in your post, it is hard to guess what might have triggered it.
To get a matrix out of your example, do this instead:
mycounts <- read.table("newMat.txt", header = T, sep = "\t",row.names=1)
mymat <- as.matrix(mycounts)
The only difference to your code is the "row.names" argument that will use the first column as row.names, so you will get something like:
AT1G01060 AT1G01170 AT1G01180
AT1G01060 0 0 0
AT1G01170 0 0 0
...
And since your mycounts object will have only one type, you can coerce it as a matrix.
Check ?read.table for more details.
HTH,
Nico
---------------------------------------------------------------
Nicolas Delhomme, PhD
Acting Manager
UPSC bioinformatics core facility
Umeå Plant Science Center,
Swedish University for Agricultural
Sciences (SLU) and Umeå University
Tel: +46 90 786 5478
Email: nicolas.delhomme@umu.se
SLU - Umeå universitet
Umeå S-901 87 Sweden
---------------------------------------------------------------
> On 30 Dec 2015, at 22:39, Fereshteh.Izadi@mpikg.mpg.de [bioc] <noreply@bioconductor.org> wrote:
>
> Activity on a post you are following on support.bioconductor.org
> User Fereshteh.Izadi@mpikg.mpg.de wrote Question: converting integer to numeric:
>
>
> hi
>
> i have a matrix but im facing this error
>
> mycounts <- read.table("newMat.txt", header = T, sep = "\t")
> head(mycounts[,1:4])
> IDs AT1G01060 AT1G01170 AT1G01180
> 1 AT1G01060 0 0 0
> 2 AT1G01170 0 0 0
> 3 AT1G01180 0 0 0
> 4 AT1G01183 0 0 0
> 5 AT1G01260 0 0 0
> 6 AT1G01380 0 0 0
>
> list object cannot be coerced to type integer
>
> how i can convert an integer to numeric please?
>
> thank you
>
>
> Post tags: software error
>
> You may reply via email or visit converting integer to numeric
>
This is not a Bioconductor question. It's about basic manipulations in R. This kind of questions is better asked on the R-help mailing list (https://stat.ethz.ch/mailman/listinfo/r-help).
You're showing us code but you omitted the most important part which is the command that actually caused the error.
The question you ask ("how i can convert an integer to numeric") doesn't seem directly related to the code you show. To convert an integer vector to a numeric vector, just call as.numeric() on it. However, your mycounts object is a data.frame, not an integer vector.
It could be that you got the error message you show us by calling as.integer() on the mycounts data.frame. A data.frame is a special kind of list hence the error message you got. Note that calling as.integer() on a list only works in the very particular case where all the list elements have a length <= 1.
Calling as.matrix() on your mycounts data.frame should work except that it will turn it into a character matrix, not an integer matrix. As Nico said, that's because one of the columns (the 1st one) in your data.frame is of type character (or factor). To get an integer matrix, you first need to get rid of this column. You can do this either by subsetting mycounts with as.matrix(mycounts[ , -1]), or by tweaking your call to read.table() as showed by Nico.
