error in density.default : x must be numeric

Question

This question might have been repeated. But even after going through the previous links, I am not able to solve this. I have a file as follows:

data <- read.table("data.txt", header=TRUE)

             Samp1           Samp2     Samp3
cg00000029  0.79015390399987 0.8301816 0.8966661
cg00000108 0.970260858767027 0.9655997 0.9699428
cg00000109 0.948456317952246 0.9209855 0.9325146
cg00000165 0.267769194351135 0.2370634 0.3867273

I wish to create a density plot out of the column (say Samp1). When I use the following

>plot(density(na.omit(data$Samp1)), col="black")

I get the following error:

Error in density.default(na.omit(data$Samp1)) : argument 'x' must be numeric

Can anyone help me know how to i correct this? I have created density plots for similar files, but did not get this error. It is only for this file.

Your help appreciated. Thanks in advance..

Answer 1

Well, for some reason your data is non-numeric: have you tried using as.numeric() to coerce it into the right type?

Edit : Using unlist() to convert it out of a list type seems to be the answer

Answer 2

I was getting the same problem, I was able to solve it by doing:

a<- density(treeDF$real)
plot(a$x, a$y, col="turquoise2")
b<- density(treeDF$rpart)
lines(b$x, b$y, col="deeppink3")
c<- density(treeDF$rpart_Fourier)
lines(c$x, c$y, col="blue")
legend("topright", legend = c("real value", "rpart()", "rpart() + Fourier tem"),
    col=c("turquoise2", "deeppink3", "blue"), lty=1:1, cex=0.8, box.lty = 0)