read.table and missing values

I'd like to share data files (tab delimited text files)
that i read in via read.table with other applications.
Missing data are empty fields (two tabs following each other)
I couldn't find a way yet to 'convince' R to interprete this to be
missing values.

It's the default ....

The read.table option na.strings="" or na.strings='' didn't work.
using any special character in the data file for missings (like '#')
and defining it for missing using na.strings="#" worked.
So I was able to solve the problem doing a global replace.
But I'd prefer to read in the data files as is and wonder whether
there is a possibility to that in R.
Data Import/Export Manual says that empty fields in numeric columns
are regarded as missing values.
But with my versions of R (Windows 1.2.3, Mac) it didn't work
Documention, FAQ and the mailing archive didn't have any further hint.
Thanks for any advice

It works as documented, and as documented does what you want (so no
further hint is needed).  For file foo.dat

a,b,c
1,2,3
4,,6
,8,9

I get

read.table("foo.dat", header=T, sep=",")

a  b c
1  1  2 3
2  4 NA 6
3 NA  8 9

as advertised.

You have set sep="\t", haven't you?  The default separator is white space,
and that will swallow two tabs.  The following had tabs in originally,

a	b	c
1	2	3
4		6
	8	9

and gives

read.table("foo.dat", header=T, sep="\t")

a  b c
1  1  2 3
2  4 NA 6
3 NA  8 9

So, if you do that it should work for you too.  And if you want to claim
on R-help that things do not work, please include an example so we can see
what you are doing (or not doing).

Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._

Thanks - everything is o.k. and working

a,b,c
1,2,3
4,,6
,8,9

I get

 read.table("foo.dat", header=T, sep=",")

   a  b c
1  1  2 3
2  4 NA 6
3 NA  8 9

as advertised.

You have set sep="\t", haven't you?  The default separator is white space,
and that will swallow two tabs.  The following had tabs in originally,

This did solve the problem. I didn't set sep="\t". R read complete
tab delimited data files nontheless and complained only when it met 
missing data.
This confused me.

a	b	c
1	2	3
4		6
8	9

and gives

 read.table("foo.dat", header=T, sep="\t")

   a  b c
1  1  2 3
2  4 NA 6
3 NA  8 9

So, if you do that it should work for you too.  And if you want to claim
on R-help that things do not work, please include an example so we can see
what you are doing (or not doing).

sorry for having missed to send an example.

Peter

---------------------------------------------------------------------
Peter Breuer, http://www.psych.uni-goettingen.de/home/breuer
Georg-Elias-Mueller-Inst. f. Psychologie, Gosslerstr. 14, 37073 
Goettingen, Germany
Tel: (+49)551-39-3588,  Fax: (+49)551-39-3662
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._

Thanks - everything is o.k. and working

[...]

You have set sep="\t", haven't you?  The default separator is white space,
and that will swallow two tabs.  The following had tabs in originally,

This did solve the problem. I didn't set sep="\t". R read complete
tab delimited data files nontheless and complained only when it met
missing data.
This confused me.

Using read.delim (same help page) might be less confusing, although I
think fill = TRUE might also cause confusion.

Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._

<snip>

You have set sep="\t", haven't you?  The default separator is white space,
and that will swallow two tabs.  The following had tabs in originally,

a	b	c
1	2	3
4		6
8	9

and gives

 read.table("foo.dat", header=T, sep="\t")

   a  b c
1  1  2 3
2  4 NA 6
3 NA  8 9

<snip>

Cool! Thanks a lot! I had the same problem a while ago and solved it
by replacing two consecutive tabs with \t""\t. Now I see what the
problem was: When I'm not using R I'm using StatView and StatView
defaults to the tab as separator. (I don't remember what it was in
SPSS or Systat.) So I assumed that it's the same with R.
I didn't notice the sep = "" in the Usage section of the help page.
To make the default value more visible in this case it would be
nice if in the Arguments/sep section for read.table it would read
... If sep = "" (the default) the separator is ...
                 -------------

Cheers,

Ragnar

(Must've said "blimey" once too often -
  now I'm really blind every now and then ;)
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._

<snip>
To make the default value more visible in this case it would be
nice if in the Arguments/sep section for read.table it would read
... If sep = "" (the default) the separator is ...
                -------------

<snip>

Oops, I meant:
... If sep = "" (the default for read.table) the separator is ...

Ragnar
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._