[R] Output of order() incorrectly ordered?

[Shirley Wu]

Thanks everyone for the input. It turns out that either read.table or  
write.table was having trouble with single quotes in the data  
(read.table has an option for specifying quote characters or ignoring  
them, but as far as I could tell, the data frame was still sorting  
correctly and the problem only appeared in the output file).

Once I removed the single quotes from my input files, everything  
worked fine.

Thanks,
Shirley


On Mar 25, 2008, at 10:03 AM, Charles C. Berry wrote:

> On Mon, 24 Mar 2008, Shirley Wu wrote:
>
>> Hello,
>>
>> I have a data frame consisting of four columns and would like to sort
>> based on the first column and then write the sorted data frame to a
>> file.
>>
>> > df <- read.table("file.txt", sep="\t")
>> where file.txt is simply a tab-delimited file containing 4 columns of
>> data (first 2 numeric, second 2 character). I then do,
>>
>> > df_ordered <- df[order(df$V1), ]
>>
>> OR, I assume equivalently,
>>
>> > df_ordered <- df[ do.call(order, df), ]
>>
>> and then,
>>
>> > write.table(df_ordered, file="newfile.txt", ...)
>>
>> The input data file looks like this:
>>
>> 0.083044        375.276 680220  majority
>> 5.50816e-09     2.48914e-05     26377   conformation
>> 0.000169618     0.766505        1546938 interaction
>> 3.90425e-05     0.176433        1655338 vitamin
>> 0.0378182       170.9   1510941 array
>> 3.00359e-07     0.00135732      69421   oligo(dT)-cellulose
>> 1.01517e-13     4.58754e-10     699918  elastase
>> ...
>>
>> I'd like the output file to look the same except sorted by the first
>> column. The output of the commands above give me something that is
>> sorted in some places but not sorted in others:
>>
>> [sorted section]
>> ...
>> 1.87276e-07     0.000846299     1142090 vitamin K
>> 1.89026e-07     0.000854207     917889  leader peptide
>> 1.90884e-07     0.000862605     31206   s
>> 0.00536062      24.2246 1706420 prevent
>> 5.42648e-05     0.245223        1513041 measured
>> 5.42648e-05     0.245223        1513040 measured
>> 0.019939        90.1044 12578   fly
>> 0.00135512      6.12377 61688   GPI
>> 0.00124421      5.62257 681915  content
>> 0.0128271       57.9655 681916  estimated
>> ...
>> [sorted section]
>> ...
>> [unsorted section]
>> ...
>> [etc]
>>
>> I'm not sure if this is a problem with the input data or with order()
>> or what. I am only doing this in R because many of my numeric values
>> are expressed in exponential notation and UNIX sort does not handle
>> this to my knowledge, but this behavior baffles me. I am pretty new
>> to R so it's possible I'm missing something.
>>
>> Any insight would be greatly appreciated!
>
> I suspect that the first column contains something that cannot be  
> rendered as a numeric value. Probably you have leading blanks  
> before some of the numbers.
>
> The result is that the first column is a factor, which will be  
> ordered according to the character collating sequence in your  
> locale after coercing it to character. (I am guessing here, but it  
> appears to do this on my PC.)
>
> Try
>
> df <- read.table("file.txt", sep="\t", strip.white=TRUE )
>
> and see if the ordering agrees with your intuition.
>
> HTH,
>
> Chuck
>
>
>>
>> Thanks,
>> -Shirley
>> graduate student
>> Stanford University
>>
>> ______________________________________________
>> 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.
>>
>
> Charles C. Berry                            (858) 534-2098
>                                             Dept of Family/ 
> Preventive Medicine
> E mailto:cberry at tajo.ucsd.edu	            UC San Diego
> http://famprevmed.ucsd.edu/faculty/cberry/  La Jolla, San Diego  
> 92093-0901
>
>