R help - Apr 2016 - Using a function with apply Error: undefined columns selected

I am trying to write a function that can be used to apply to process all the
columns of a data.frame. If you will run the code below, you will get the error
message undefined columns selected. I hope someone will be able to teach me what
I am doing wrong.
Thank you,
John 
 
# create data frame.
guppy
 
fract2 <- function(col,data) {
  cat("Prove we have passed the data frame\n")
  print(data)
  
  # Get the name of the column being processed.
  zz<-deparse(substitute(col))
  cat("Column being processed\n")
  print(zz)
  p<-sum(data[,zz]!="")/length(data[,zz])
  return(p)
}
 
apply(guppy,2,fract2,data=guppy)
John David Sorkin M.D., Ph.D.
Professor of Medicine
Chief, Biostatistics and Informatics
University of Maryland School of Medicine Division of Gerontology and Geriatric
Medicine
Baltimore VA Medical Center
10 North Greene Street
GRECC (BT/18/GR)
Baltimore, MD 21201-1524
(Phone) 410-605-7119
(Fax) 410-605-7913 (Please call phone number above prior to faxing) 

Confidentiality Statement:
This email message, including any attachments, is for the sole use of the
intended recipient(s) and may contain confidential and privileged information.
Any unauthorized use, disclosure or distribution is prohibited. If you are not
the intended recipient, please contact the sender by reply email and destroy all
copies of the original message.

Hi John,
First, apply isn't guaranteed to work on data frames. There are two
easy ways to do something like this, but we had better have a data
frame:

guppy<-data.frame(taste=rnorm(10,5),
 crunch=rnorm(10,5),satiety=rnorm(10,5))

If you just want to apply a function to all or a subset of columns of
a data frame, a for loop can be used:

fract2.1<-function(col,data) {
 p<-sum(data[,col],na.rm=TRUE)/sum(!is.na(data[,col]))
 return(p)
}
for(col in 1:ncol(guppy)) print(fract2.1(col,guppy))

If you really do want to use an "*apply" function, then the function
has to be written for each column, not the entire data frame:

fract2.2<-function(x) return(sum(x,na.rm=TRUE)/sum(!is.na(x)))
sapply(guppy,fract2.2)

and if you want a subset of the columns, you will have to do it before
you let sapply get into it.

Jim



On Fri, Apr 8, 2016 at 8:39 AM, John Sorkin <jsorkin at
grecc.umaryland.edu> wrote:
> I am trying to write a function that can be used to apply to process all
the columns of a data.frame. If you will run the code below, you will get the
error message undefined columns selected. I hope someone will be able to teach
me what I am doing wrong.
> Thank you,
> John
>
> # create data frame.
> guppy
>
> fract2 <- function(col,data) {
>   cat("Prove we have passed the data frame\n")
>   print(data)
>
>   # Get the name of the column being processed.
>   zz<-deparse(substitute(col))
>   cat("Column being processed\n")
>   print(zz)
>   p<-sum(data[,zz]!="")/length(data[,zz])
>   return(p)
> }
>
> apply(guppy,2,fract2,data=guppy)
> John David Sorkin M.D., Ph.D.
> Professor of Medicine
> Chief, Biostatistics and Informatics
> University of Maryland School of Medicine Division of Gerontology and
Geriatric Medicine
> Baltimore VA Medical Center
> 10 North Greene Street
> GRECC (BT/18/GR)
> Baltimore, MD 21201-1524
> (Phone) 410-605-7119
> (Fax) 410-605-7913 (Please call phone number above prior to faxing)
>
> Confidentiality Statement:
> This email message, including any attachments, is for ...{{dropped:12}}

> On Apr 7, 2016, at 3:39 PM, John Sorkin <jsorkin at
grecc.umaryland.edu> wrote:
> 
> I am trying to write a function that can be used to apply to process all
the columns of a data.frame. If you will run the code below, you will get the
error message undefined columns selected. I hope someone will be able to teach
me what I am doing wrong.
> Thank you,
> John 
> 
> # create data frame.
> guppy
> 
> fract2 <- function(col,data) {
>  cat("Prove we have passed the data frame\n")
>  print(data)
> 
>  # Get the name of the column being processed.
>  zz<-deparse(substitute(col))
>  cat("Column being processed\n")
>  print(zz)
>  p<-sum(data[,zz]!="")/length(data[,zz])
>  return(p)
> }
> 
> apply(guppy,2,fract2,data=guppy)
At the point where the error is about to occur during the first column being
processed, this is what had been printed:

Column being processed
[1] "newX[, i]"

So it should be no surprise that the actual error was:

Error in `[.data.frame`(data, , zz) : undefined columns selected

It occurred at one of the two points where you tried `data[,zz]`

You need to pass colnames(guppy) to fract2 and work with the character values.

All of the *apply function pass only values so they do not pass the column names
for testing. A possible exception might be the colnames of a vector being
avaialble when using apply with an index of 1.

-- 


> John David Sorkin M.D., Ph.D.
> 
Snipped

-- 
David Winsemius
Alameda, CA, USA