R help - Aug 2006 - getting sapply to skip columns with non-numeric data?

getting s-apply to skip columns with non-numeric data?
I have a dataframe ?x? of w columns.

Some columns are numeric, some are not.

I wish to create a function to calculate the mean and
standard deviation of each numeric column, and then
?bind? the column mean and standard deviation to the
bottom of the dataframe.

e.g. 

tempmean <- apply(data.frame(x), 2, mean, na.rm = T)
xnew <- rbind(x,tempmean)

I am running into one small problem
what is the best
way to have sapply ?skip? the non-numeric data and
return NA?s?

Use the first few rows of iris as test data and try this
where isnum is 1 for each numeric column and NA for
others.

irish <- head(iris)
isnum <- ifelse(sapply(iris, class) == "numeric", 1, NA)
iris.data <- data.matrix(iris)
rbind(iris, colMeans(iris.data) * isnum, sd(iris.data) * isnum)


On 8/17/06, r user <ruser2006 at yahoo.com> wrote:
> getting s-apply to skip columns with non-numeric data?
> I have a dataframe "x" of w columns.
>
> Some columns are numeric, some are not.
>
> I wish to create a function to calculate the mean and
> standard deviation of each numeric column, and then
> "bind" the column mean and standard deviation to the
> bottom of the dataframe.
>
> e.g.
>
> tempmean <- apply(data.frame(x), 2, mean, na.rm = T)
> xnew <- rbind(x,tempmean)
>
> I am running into one small problem?what is the best
> way to have sapply "skip" the non-numeric data and
> return NA's?
>
> ______________________________________________
> R-help at stat.math.ethz.ch 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.
>

There's something that either you have not thought of or neglected to tell
us:  If you have k variables in the data frame, you will need a data frame
of k variables and one row to be able to rbind() to the bottom of the
original one.  What are you going to put in place for non-numeric variables?

Perhaps this might help:

R> dat <- data.frame(f=factor(1:3), x=3:5, y=6:4)
R> rbind(dat, as.data.frame(lapply(dat, function(x) if (!is.numeric(x)) NA
else mean(x))))
      f x y
1     1 3 6
2     2 4 5
3     3 5 4
11 <NA> 4 5

Andy 

From: r user
> 
> getting s-apply to skip columns with non-numeric data?
> I have a dataframe "x" of w columns.
> 
> Some columns are numeric, some are not.
> 
> I wish to create a function to calculate the mean and 
> standard deviation of each numeric column, and then "bind" 
> the column mean and standard deviation to the bottom of the dataframe.
> 
> e.g. 
> 
> tempmean <- apply(data.frame(x), 2, mean, na.rm = T) xnew <- 
> rbind(x,tempmean)
> 
> I am running into one small problem...what is the best way to 
> have sapply "skip" the non-numeric data and return NA's?
> 
> ______________________________________________
> R-help at stat.math.ethz.ch 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.
> 
>