Reclassify string values

Hi All,

Is there a simple way to convert a string such as c("A", "B" ,"C", "D") to a
string of c("Group1", "Group1", "Group2", "Group2"). Naturally I could use
the factor function as below but I don't like seeing that warning message
(and I don't want to turn off warning messages). Perhaps a function called
"reclassify" or "recategorize"?

Zev

x<-LETTERS[1:4]
x2<-as.character(factor(x, levels=LETTERS[1:4], labels=rep(c("Group1",
"Group2"), each=2)))

Warning message:
In `levels<-`(`*tmp*`, value = c("Group1", "Group1", "Group2", "Group2" :
?duplicated levels will not be allowed in factors anymore

translation

On Thu, Nov 3, 2011 at 11:59 AM, Zev Ross<zev at zevross.com>  wrote:

Hi All,

Is there a simple way to convert a string such as c("A", "B" ,"C", "D") to a
string of c("Group1", "Group1", "Group2", "Group2"). Naturally I could use
the factor function as below but I don't like seeing that warning message
(and I don't want to turn off warning messages). Perhaps a function called
"reclassify" or "recategorize"?

Zev

x<-LETTERS[1:4]
x2<-as.character(factor(x, levels=LETTERS[1:4], labels=rep(c("Group1",
"Group2"), each=2)))

Warning message:
In `levels<-`(`*tmp*`, value = c("Group1", "Group1", "Group2", "Group2" :
  duplicated levels will not be allowed in factors anymore

If you want to "translate", why not first build a translation table

tt = cbind(LETTERS[1:4], c("group1", "group1", "group2", "group2"))

then apply it on an example:

xx = sample(LETTERS[1:4], 20, replace = TRUE)

translation = tt[ match(xx, tt[, 1]), 2]

translation

  [1] "group2" "group2" "group2" "group2" "group2" "group1" "group2" "group1"
  [9] "group2" "group1" "group1" "group2" "group2" "group2" "group1" "group2"
[17] "group2" "group1" "group1" "group2"

Or did I misunderstand your intent?

Peter

Hi Peter,

Thanks for the response. What you've suggested works fine but I'm looking
for something that is simpler than my solution and avoids the pesky warning
message. Your response avoids the warning message but just as complex (if
not more). I just assumed there would be a function along the lines of:

mydata <- c("A", "C", "A", "D", "B", "B")
reclassify(mydata, inCategories=c("A", "B" ,"C", "D"),
?outCategories=c("Group1", "Group1", "Group2", "Group2"))

[1] "Group1" "Group2" "Group1" "Group2" "Group1" "Group1"

On Thu, Nov 3, 2011 at 1:31 PM, Zev Ross <zev at zevross.com> wrote:

Hi Peter,

Thanks for the response. What you've suggested works fine but I'm  
looking
for something that is simpler than my solution and avoids the pesky  
warning
message. Your response avoids the warning message but just as  
complex (if
not more). I just assumed there would be a function along the lines  
of:

mydata <- c("A", "C", "A", "D", "B", "B")
reclassify(mydata, inCategories=c("A", "B" ,"C", "D"),
 outCategories=c("Group1", "Group1", "Group2", "Group2"))

[1] "Group1" "Group2" "Group1" "Group2" "Group1" "Group1"

But of course, except sometimes you have to write the function  
yourself.

reclassify = function(data, inCategories, outCategories)
{
  outCategories[ match(data, inCategories)]
}

Sorry I can't make it any simpler than a 1-line solution :)