R help - Oct 2011 - question regarding intersect function

Hi

 

I have probably a very simple question but I'm going crazy trying to find
the solution.

 

I have two data.frames with headers and I'm doing an intersection between
them by names, such that the intersected data.frames are returned by:
df1[intersect(names (df1), names(df2))] and the same for df2

 

Now, I want to have all the opposite data that did not intersect. I tried to
do: df1[-intersect(names (df1), names(df2))]  and some other stuff but it
didn't work.

 

Any ideas?

 

Thanks, S.B.


	[[alternative HTML version deleted]]

On Oct 25, 2011, at 6:07 AM, Samir Benzerfa wrote:
> Hi
>
>
>
> I have probably a very simple question but I'm going crazy trying to  
> find
> the solution.
>
>
>
> I have two data.frames with headers and I'm doing an intersection  
> between
> them by names, such that the intersected data.frames are returned by:
> df1[intersect(names (df1), names(df2))] and the same for df2
>
>
>
> Now, I want to have all the opposite data that did not intersect. I  
> tried to
> do: df1[-intersect(names (df1), names(df2))]  and some other stuff  
> but it
> didn't work.
Right. You cannot use negative indexing with character column  
identifiers.

Two methods you can try:
df1[ setdiff( names(df1), names(df2)) ]
df2[ setdiff( names(df2), names(df1)) ]

Or:
df1[!names(df1) %in% names(df2)]

(They both use match() "under the hood" so there would also be a  
match() based solution as well.)

--

David Winsemius, MD
West Hartford, CT