[R] Reading data with blank elements

Marc Schwartz (via MN) mschwartz at mn.rr.com
Fri Jul 21 18:08:51 CEST 2006

On Fri, 2006-07-21 at 05:43 -0700, Ahamarshan jn wrote:
> Hi,
>  I have a dataset saved in *.csv format, that contains
> 13 columns (the first column being the title name and
> the rest experiments) and about 2500 rows.
> Not all columns in the row have data in it
> i.e for eg
> BS00,-0.084,0.0136,-0.1569,-0.6484,1.103,1.7859,0.40287,0.5368,0.08461,-0.1935,-0.147974,0.30685
> BS01,0.491270283,0.875826172,,,,,,,,,,
> BS02,0.090794476,0.225858954,,,0.32643,0.34317,0.133145295,,,0.115832599,0.47636458,
> BS03,0.019828221,-0.095735935,-0.122767219,-0.0676,0.002533,-0.1510361,0.736247,2.053192,-0.423658,0.4591219,1.1245015,
> BS04,-0.435189342,-0.041595955,-0.781281128,-1.923036,-3.230167102,,,,0.152322609,-1.495513519,,
> I am using R to perform a correlation, but I am
> getting an error while trying to read the data as
> ">
> person.data<-read.table("datafile.csv",header=TRUE,sep=',',row.names=1)
> Error in scan (file = file, what = what, sep = sep,
> quote = quote, dec = dec,  : 
>         line 1919 did not have 13 elements
> Execution halted "
> The error looks as though there is a problem with the
> last element being not read when it is blank. I could
> introduce terms like "na" to the blank elements but I
> donot want to do that because this will hinder my
> future analysis. 
> Can some one suggest me a solution to overcome this
> problem while reading the data? , or is there
> something that I have missed to make the data
> readable. 
> Thank you in advance, 
> PS: The data was imported from a experiment and saved
> in excel sheet as a *.csv and then used.

You have already had other replies, to which I would add, be sure to
read Chapter 8 in the R Import/Export Manual regarding importing Excel
files and other options besides exporting to a CSV file.

In addition, the issue of Excel generating CSV files with the last
column missing on some rows is a known issue and is reported in the MSKB


Even though the latest version of Excel listed in the article as being
relevant is 97, I had this problem with 2000 and 2003 as well.

I would instead use OpenOffice.org's Calc to do the export when this was
required. Calc did not have this problem.


Marc Schwartz

More information about the R-help mailing list