[R] subsetting with condition

[David Winsemius]

On Jun 1, 2011, at 7:00 PM, kristina p wrote:

> Dear R Team,
>
> I am a new R user and I am currently trying to subset my data under a
> special condition. I have went through several pages of the subsetting
> section here on the forum, but I was not able to find an answer.
>
> My data is as follows:
>
> ID                      NAME       MS     Pol. Party
> 1                           John       x       F
> 2                           Mary       s       S
> 3                           Katie      x       O
> 4                           Sarah      p       L
> 5                           Martin      x      O
> 6                           Angelika   x      F
> 7                            Smith      x      O
> ....

Assume this is in a dataframe, 'pol', and that you have corrected the  
error in colnames, so that it is Pol_Party. the ave function is  
particularly useful when you need to have a vector that "lines up  
along side" the other columns

pol[ave(seq_along(pol$ID), pol$Pol_Party, FUN=length) >= 3 , ]
   ID   NAME MS Pol_Party
3  3  Katie  x         O
5  5 Martin  x         O
7  7  Smith  x         O

(The use of seq_along ensures you will get duplicates of ID that are  
in any qualifying Parties.

Another way to generate the values would be to table()-ulate and pick  
out the names of qualifying Parties:

 > pol[ pol$Pol_Party %in% names(tabl.party)[tabl.party >= 3], ]
   ID   NAME MS Pol_Party
3  3  Katie  x         O
5  5 Martin  x         O
7  7  Smith  x         O


> I am intested in only those observations, where there are at least  
> three
> members of 1 political party. That is, I need to throw out all cases  
> in the
> example above, except for members of party "O".

Both methods use logical indexing with the "[.data.frame" function,

>
> Would really appreciate your help.

-- 
David Winsemius, MD
West Hartford, CT