[R] unique

Milan Bouchet-Valat nalimilan at club.fr
Tue Oct 16 15:48:39 CEST 2012

Le mardi 16 octobre 2012 à 14:45 +0200, paladini a écrit :
> Hello everybody,
> I've got a problem concerning the function unique. I have got a 
> data.frame "shopdata" with 1000 shop which were evaluated at different 
> points in time.
> With function subset I chose those shops with more then 10 employee and 
> store it in data.frame "bigshopdata" with 700 shops.
>   bigshopdata=subset(shopdata, shopdata$employee>10)
> Now I use unique(bigshopdata$name) to ensure that each shop name is 
> only listed  ones and not two or three times because of an evaluation at 
> different dates.
> What happens is that unique eliminates also those shops names which 
> appear only ones in bigshopdata but twice or more often in shopdata.
> But that is not what I want to. I'm only interessted in multiple 
> appearance in bigshopdata not in an possible multible appearance in the 
> original data.farme.
> How can I use unique to get what I want? Or is there an alternative 
> function?
Please provide a reproducible example with a subset of your data, or
with sample data. Your problem does not make sense to me. In particular,
unique() does not "eliminate" as you say, it retains all values that
appear at least once; so a value appearing twice or more will not be


More information about the R-help mailing list