[R] VLOOKUP in R - tried everything.

[David Winsemius]

On Jun 3, 2011, at 12:43 PM, peter dalgaard wrote:

>
> On Jun 3, 2011, at 16:59 , bjmjarrett wrote:
>
>> I am attempting to emulate the VLOOKUP function from Excel in R.
>>
>> I want to compare one column (coll.minus.release) with another
>> (release.days) to get the number of parasitoid released at that time
>> (TotalParasitoids).
>>
>> for example:
>>
>> coll.minus.release	release.days	ParasitoidTotal
>> -12	                             -266	             1700
>> 8	                             -259	             1000
>> 8	                             -225	             1000
>> 28	                             -216	             1000
>> 41	                             -28	             1148
>> 77	                              -12	             1144
>> 105	                               0	             1160
>> 105	                               8	              972
>> 125	                              28	             1146
>> 125	                              41	             1004
>> 125	                              77	             1003
>> 125	                              97	             1010
>> ....
>> 2772	                              NA	               NA
>> 2801	                              NA	               NA
>> 2834	                              NA	               NA
>>
>>
>> vlookup <- function(x) data[data$release.days==x,6] # as I have  
>> three other
>> columns that are not of interest
>>
>> vlookup(-12) = 1144, and so on, which is great.
>>
>> However, when I try:
>>
>> unlist(sapply(coll.minus.release,vlookup)) to apply it to the whole
>> coll.minus.release
>>
>> it works up to a point, as it doesn't give me 132 values for the  
>> 132 values
>> of coll.minus.release. Is this because the table of release.days and
>> TotalParasitoid has less values than coll.minus.release (108  
>> compared to
>> 132)? To fill the gap I put in 0, and as none of the  
>> coll.minus.release
>> values = 0 I think it wouldn't affect it.
>
>
> I wager that a look at setdiff(coll.minus.release,release.days) and  
> vice versa would be illuminating. Notice that with your definition,  
> vlookup(31415926) or any other number absent from release.days gives  
> a zero-length vector.
>
> Presumably, you are looking for match().

If I remember correctly VLOOKUP (and HLOOKUP)  require sorted  
cutpoints and picks the lowest number that the vector element exceeds  
or equals. If so, then I suspect that an indexing strategy with  
findInterval should work correctly if the dataframe is sorted with  
order(). If you post a cut-pasteable example with desited input and  
output, I sure someone can rig something up.

-- 
David.

>
>
>>
>> Other things I have tried include findInterval and match.
>>
>> data[findInterval(x=data$coll.minus.release,vec=data 
>> $release.days,"ParasitoidTotal")]
>>
>> didn't work as it said vec must be sorted non-decreasingly and  
>> didn't work
>> when I randomised the release.days and ParasitoidTotal columns as  
>> it doesn't
>> matter which order they are in.
>>
>> Thanks for reading all the way through - I wanted all the  
>> information I felt
>> you might need to help me in it.
>>
>> Any help will be greatly appreciated.
>>
>> --
>> View this message in context: http://r.789695.n4.nabble.com/VLOOKUP-in-R-tried-everything-tp3571107p3571107.html
>> Sent from the R help mailing list archive at Nabble.com.
>>
>> ______________________________________________
>> R-help at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
>
> -- 
> Peter Dalgaard
> Center for Statistics, Copenhagen Business School
> Solbjerg Plads 3, 2000 Frederiksberg, Denmark
> Phone: (+45)38153501
> Email: pd.mes at cbs.dk  Priv: PDalgd at gmail.com
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.

David Winsemius, MD
West Hartford, CT