R help - Jan 2016 - Trying to use name of difference variable created in a function in call to wilcox.test function.

I am trying to write a function which will allow me to analyze many variables
all of which have the same form, Base.stem and Disch.stem, where stem varies
from variable to variable, e.g. Base.rolling and Disch.rolling, Base.standing,
Disch.standing.
 
I want to pass a dataframe and the stem of the name of the pre and post
intervention variables. In my function I create the full pre and post
intervention Base.stem and Disch.stem variables names (by using the paste
function, varpre=paste("Base.",stem,sep="") and
varpost=paste("Disch.",stem,sep="")). I also create and add
to the dataframe the change variable (i.e. post-pre) formed as
ch<-paste("Change.",stem,sep="") and compute the change
in the variable:    data[,ch]=data[,varpo]-data[,varpre].
 
If, for example, I run the function doit(rolling, data), the function creates
the full pre-intervention variable name Base.rolling, the post-intervention
variable name Disch.rolling, and computes Change.rolling =
Disch.rolling-Base.rolling. I want to use Change.rolling in a function, but I
don't want to have to expressly specify the name of the outcome variable,
i.e.
 
 
fitwilcox<-wilcox.test(Change.rolling~RANDOMIZED,data=data,na.action=na.omit)
 
I want to specify the constructed variable, i.e.
 
 fitwilcox<-wilcox.test(ch~RANDOMIZED,data=data,na.action=na.omit)
 
This does not work because ch is not in the dataframe, but Change.rolling is.
How can I modify the call to the Wilcox.test function so I can specify the
change variable constructor. I tried the following but it does not work:
 
 fitwilcox<-wilcox.test(eval(ch,
parent.frame())~RANDOMIZED,data=data,na.action=na.omit)
 
DATA:
 
> dput(data2)structure(list(Subj = c(115, 116, 117, 118, 119, 120, 121, 122,
123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138,
139, 140, 141, 142, 143, 144, 145, 146, 147),     RANDOMIZED = c(1, 1, 2, 1, 2,
1, 2, 1, 1, 2, 1, 2, 1, 2,     1, 1, 2, 2, 1, 2, 1, 1, 2, 2, 1, 2, 1, 2, 2, 2,
2, 2, 2),     Base.rolling = c(0, 6, 6, 3, 3, 1, 4, 6, 4, 6, 4, 1, 3, NA,     1,
2, 4, 1, 4, 4, 4, 1, 5, 3, 2, 3, 1, 4, 0, 4, 1, NA, 3),     Disch.rolling = c(6,
NA, NA, 6, 6, 3, 6, 6, 6, NA, NA, 6,     1, NA, 6, 6, 6, NA, 6, 6, 6, NA, 4, NA,
6, 6, NA, 6, NA,     6, 6, NA, NA)), .Names = c("Subj",
"RANDOMIZED", "Base.rolling", "Disch.rolling"),
class = "data.frame", row.names = c(NA, -33L))
 
CODE:

doit <- function(var,data){
  var <-    deparse(substitute(var))
  print(var)
  # Create name for basline variable.
  varpre<-paste("Base.",var,sep="")
  # Create name for post variable.
  varpo<-paste("Disch.",var,sep="")
  # Create a name for the change variable.
  ch<-paste("Change.",var,sep="")
  print(ch)
  # Compute change.
  data[,ch]=data[,varpo]-data[,varpre]
  
  cat("\nData used in the analyses\n")
  print(data[,c("Subj","RANDOMIZED",ch)])
  # This works when I expressly specify the change variable.
  cat("This works, expressly specify change variable\n")
 
fitwilcox1<-wilcox.test(Change.rolling~RANDOMIZED,data=data,na.action=na.omit)
  print(fitwilcox1)
  
  # This does not work.
  cat("This does NOT work, try to use created change variable.\n")
  fitwilcox2<-wilcox.test(eval(ch,
parent.frame())~RANDOMIZED,data=data,na.action=na.omit)
  print(fitwilcox2)
  # Compute wilcoxon statistic.
}
doit(rolling,data2)
 
 
Thank you,
John
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.

Is the following function, myFormula(), what you are looking for?

myFormula <- function(stem, env = parent.frame()) {
    eval(bquote(.(as.name(paste0("Disch.",stem))) ~
.(as.name(paste0("Base.",stem))),
list(stem=as.name("rolling"))), envir=env)
}
str(myFormula("myStem"))
#Class "formula" language Disch.rolling ~ Base.rolling
# - attr(*, ".Environment")= <environment: R_GlobalEnv>


Bill Dunlap
TIBCO Software
wdunlap tibco.com

On Tue, Jan 12, 2016 at 3:10 PM, John Sorkin <jsorkin at
grecc.umaryland.edu>
wrote:
> I am trying to write a function which will allow me to analyze many
> variables all of which have the same form, Base.stem and Disch.stem, where
> stem varies from variable to variable, e.g. Base.rolling and Disch.rolling,
> Base.standing, Disch.standing.
>
> I want to pass a dataframe and the stem of the name of the pre and post
> intervention variables. In my function I create the full pre and post
> intervention Base.stem and Disch.stem variables names (by using the paste
> function, varpre=paste("Base.",stem,sep="") and
> varpost=paste("Disch.",stem,sep="")). I also create and
add to the
> dataframe the change variable (i.e. post-pre) formed as
> ch<-paste("Change.",stem,sep="") and compute the
change in the variable:
> data[,ch]=data[,varpo]-data[,varpre].
>
> If, for example, I run the function doit(rolling, data), the function
> creates the full pre-intervention variable name Base.rolling, the
> post-intervention variable name Disch.rolling, and computes Change.rolling
> = Disch.rolling-Base.rolling. I want to use Change.rolling in a function,
> but I don't want to have to expressly specify the name of the outcome
> variable, i.e.
>
>
>
fitwilcox<-wilcox.test(Change.rolling~RANDOMIZED,data=data,na.action=na.omit)
>
> I want to specify the constructed variable, i.e.
>
>  fitwilcox<-wilcox.test(ch~RANDOMIZED,data=data,na.action=na.omit)
>
> This does not work because ch is not in the dataframe, but Change.rolling
> is. How can I modify the call to the Wilcox.test function so I can specify
> the change variable constructor. I tried the following but it does not
work:
>
>  fitwilcox<-wilcox.test(eval(ch,
> parent.frame())~RANDOMIZED,data=data,na.action=na.omit)
>
> DATA:
>
> > dput(data2)structure(list(Subj = c(115, 116, 117, 118, 119, 120, 121,
> 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136,
> 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147),     RANDOMIZED >
c(1, 1, 2, 1, 2, 1, 2, 1, 1, 2, 1, 2, 1, 2,     1, 1, 2, 2, 1, 2, 1, 1, 2,
> 2, 1, 2, 1, 2, 2, 2, 2, 2, 2),     Base.rolling = c(0, 6, 6, 3, 3, 1, 4, 6,
> 4, 6, 4, 1, 3, NA,     1, 2, 4, 1, 4, 4, 4, 1, 5, 3, 2, 3, 1, 4, 0, 4, 1,
> NA, 3),     Disch.rolling = c(6, NA, NA, 6, 6, 3, 6, 6, 6, NA, NA, 6,
>  1, NA, 6, 6, 6, NA, 6, 6, 6, NA, 4, NA, 6, 6, NA, 6, NA,     6, 6, NA,
> NA)), .Names = c("Subj", "RANDOMIZED",
"Base.rolling", "Disch.rolling"),
> class = "data.frame", row.names = c(NA, -33L))
>
>
> CODE:
>
> doit <- function(var,data){
>   var <-    deparse(substitute(var))
>   print(var)
>   # Create name for basline variable.
>   varpre<-paste("Base.",var,sep="")
>   # Create name for post variable.
>   varpo<-paste("Disch.",var,sep="")
>   # Create a name for the change variable.
>   ch<-paste("Change.",var,sep="")
>   print(ch)
>   # Compute change.
>   data[,ch]=data[,varpo]-data[,varpre]
>
>   cat("\nData used in the analyses\n")
>   print(data[,c("Subj","RANDOMIZED",ch)])
>   # This works when I expressly specify the change variable.
>   cat("This works, expressly specify change variable\n")
>
>
fitwilcox1<-wilcox.test(Change.rolling~RANDOMIZED,data=data,na.action=na.omit)
>   print(fitwilcox1)
>
>   # This does not work.
>   cat("This does NOT work, try to use created change
variable.\n")
>   fitwilcox2<-wilcox.test(eval(ch,
> parent.frame())~RANDOMIZED,data=data,na.action=na.omit)
>   print(fitwilcox2)
>   # Compute wilcoxon statistic.
> }
> doit(rolling,data2)
>
>
> Thank you,
> John
> 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:16}}

I think you're looking for ?bquote .

Something like this should give you the idea:

nm <- "yvar"
bquote( .(y)~x, list(y = as.symbol(nm )))

The key is distinguishing between the character string "yvar" and the
(language) object, as.symbol("yvar").

Cheers,
Bert
Bert Gunter

"The trouble with having an open mind is that people keep coming along
and sticking things into it."
-- Opus (aka Berkeley Breathed in his "Bloom County" comic strip )


On Tue, Jan 12, 2016 at 3:10 PM, John Sorkin
<jsorkin at grecc.umaryland.edu> wrote:
> I am trying to write a function which will allow me to analyze many
variables all of which have the same form, Base.stem and Disch.stem, where stem
varies from variable to variable, e.g. Base.rolling and Disch.rolling,
Base.standing, Disch.standing.
>
> I want to pass a dataframe and the stem of the name of the pre and post
intervention variables. In my function I create the full pre and post
intervention Base.stem and Disch.stem variables names (by using the paste
function, varpre=paste("Base.",stem,sep="") and
varpost=paste("Disch.",stem,sep="")). I also create and add
to the dataframe the change variable (i.e. post-pre) formed as
ch<-paste("Change.",stem,sep="") and compute the change
in the variable:    data[,ch]=data[,varpo]-data[,varpre].
>
> If, for example, I run the function doit(rolling, data), the function
creates the full pre-intervention variable name Base.rolling, the
post-intervention variable name Disch.rolling, and computes Change.rolling =
Disch.rolling-Base.rolling. I want to use Change.rolling in a function, but I
don't want to have to expressly specify the name of the outcome variable,
i.e.
>
>  
fitwilcox<-wilcox.test(Change.rolling~RANDOMIZED,data=data,na.action=na.omit)
>
> I want to specify the constructed variable, i.e.
>
>  fitwilcox<-wilcox.test(ch~RANDOMIZED,data=data,na.action=na.omit)
>
> This does not work because ch is not in the dataframe, but Change.rolling
is. How can I modify the call to the Wilcox.test function so I can specify the
change variable constructor. I tried the following but it does not work:
>
>  fitwilcox<-wilcox.test(eval(ch,
parent.frame())~RANDOMIZED,data=data,na.action=na.omit)
>
> DATA:
>
>> dput(data2)structure(list(Subj = c(115, 116, 117, 118, 119, 120, 121,
122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137,
138, 139, 140, 141, 142, 143, 144, 145, 146, 147),     RANDOMIZED = c(1, 1, 2,
1, 2, 1, 2, 1, 1, 2, 1, 2, 1, 2,     1, 1, 2, 2, 1, 2, 1, 1, 2, 2, 1, 2, 1, 2,
2, 2, 2, 2, 2),     Base.rolling = c(0, 6, 6, 3, 3, 1, 4, 6, 4, 6, 4, 1, 3, NA, 
1, 2, 4, 1, 4, 4, 4, 1, 5, 3, 2, 3, 1, 4, 0, 4, 1, NA, 3),     Disch.rolling =
c(6, NA, NA, 6, 6, 3, 6, 6, 6, NA, NA, 6,     1, NA, 6, 6, 6, NA, 6, 6, 6, NA,
4, NA, 6, 6, NA, 6, NA,     6, 6, NA, NA)), .Names = c("Subj",
"RANDOMIZED", "Base.rolling", "Disch.rolling"),
class = "data.frame", row.names = c(NA, -33L))
>
>
> CODE:
>
> doit <- function(var,data){
>   var <-    deparse(substitute(var))
>   print(var)
>   # Create name for basline variable.
>   varpre<-paste("Base.",var,sep="")
>   # Create name for post variable.
>   varpo<-paste("Disch.",var,sep="")
>   # Create a name for the change variable.
>   ch<-paste("Change.",var,sep="")
>   print(ch)
>   # Compute change.
>   data[,ch]=data[,varpo]-data[,varpre]
>
>   cat("\nData used in the analyses\n")
>   print(data[,c("Subj","RANDOMIZED",ch)])
>   # This works when I expressly specify the change variable.
>   cat("This works, expressly specify change variable\n")
>  
fitwilcox1<-wilcox.test(Change.rolling~RANDOMIZED,data=data,na.action=na.omit)
>   print(fitwilcox1)
>
>   # This does not work.
>   cat("This does NOT work, try to use created change
variable.\n")
>   fitwilcox2<-wilcox.test(eval(ch,
parent.frame())~RANDOMIZED,data=data,na.action=na.omit)
>   print(fitwilcox2)
>   # Compute wilcoxon statistic.
> }
> doit(rolling,data2)
>
>
> Thank you,
> John
> 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}}