[R] alternative to matching/merge?

hadley wickham h.wickham at gmail.com
Fri Jun 13 23:53:53 CEST 2008


On Fri, Jun 13, 2008 at 11:45 AM, jim holtman <jholtman at gmail.com> wrote:
> What is the structure of 'd.frame' and 'segFile'?  Run Rprof so that
> we can see which of the functions it is spending its time in.  What
> happens if x$index is not in seqFile$index?  Are the values in the
> 'index' unique in both structures?  Subsetting a data frame can be
> expensive when compared to using a matrix.  Could you use a matrix
> instead of a data frame; are all the columns the same mode?  Again
> either a subset of data would be helpful or an 'str' on the data
> objects being used so that we can understand what they are.

A few other ideas to try:

 * try merging do.call("rbind", d.frame) and seqFile, and then
spliting the results back up

 * try turning giving seqFile rownames (rownames(seqFile) <-
seqFile$index) and then use character matching:  cbind(x, seqFile[
as.character(x$index)]

 * if there is a one to one corresponding between index in seqFile and
all data.frames in d.frame, merge all of the d.frames together, order
both by index then just cbind

Hadley


-- 
http://had.co.nz/



More information about the R-help mailing list