[R] selecting rows with more than x occurrences in a given column (data type is names)

Marc Schwartz marc_schwartz at comcast.net
Tue Mar 13 16:32:22 CET 2007


On Tue, 2007-03-13 at 10:38 -0400, Mike Jasper wrote:
> Despite a long search on the archives, I couldn't find how to do this.
> Thanks in advance for what is likely a simple issue.
> 
> I have a data set where the first column is name (i.e., 'Joe Smith',
> 'Jane Doe', etc). The following columns are data associated with that
> person. I have many people with multiple rows. What I want is to get a
> new data frame out with only the people who have more than x
> occurrences in the first column.
> 
> Here's what I've done, that's not working:
> 
> Let's call my old data.frame "all.data"
> 
> table(all.data$names)>10
> 
> I get a list of names and TRUE/FALSE values. I then want to make a
> list of the TRUEs and pass that to some subset type command like
> 
> dup.names=table(all.data$names)>10
> 
> new.data=(all.data[all.data$names==dup.names,])
> 
> That's not working because the dimensions are wrong (I think). But
> even when I tried to do part of it manually (to troubleshoot) like
> this
> 
> dup.names=c('Joe Smith','Jane Doe','etc')
> 
> I got warnings and it didn't work correctly. There must be a simple
> way to do this that I'm just not seeing. Thanks.


Something like this should work:

  NewDF <- subset(all.data, names %in% unique(names[duplicated(names)]))

See ?duplicated, ?unique and ?"%in%" for more information.

HTH,

Marc Schwartz



More information about the R-help mailing list