[R] read.table: deciding automatically between two colClasses values

[Joshua Wiley]

Hi Oliver,

Look at ?readLines

I imagine something like:

tmp <- readLines(filename, n = 1L)
(do stuff with the first line to decide)
IntN <- 6 (or 4)
NumN <- 8 (or whatever)
E <- read.table(file = filename, header = TRUE, colClasses =
  c(rep("integer", IntN), "numeric", "integer", rep("numeric", NumN)), ...)

Cheers,

Josh

On Sun, Aug 28, 2011 at 7:13 AM, Oliver Kullmann
<O.Kullmann at swansea.ac.uk> wrote:
> Hello,
>
> I have a function for reading a data-frame from a file, which contains
>
>  E = read.table(file = filename,
>        header = T,
>        colClasses = c(rep("integer",6),"numeric","integer",rep("numeric",8)),
>        ...)
>
> Now a small variation arose, where
>
> colClasses = c(rep("integer",4),"numeric","integer",rep("numeric",8))
>
> needed to be used (so just a small change).
> I want to have it convenient for the user, so no user intervention shall
> be needed, but the function should choose between the two different values
> "4" and "6" here according to the header-line.
>
> Now this seems to be a problem: I found only count.fields, which
> however is not able just to read the first line. Reading the
> whole file (just to read the first line) is awkward, and also these
> files typically have millions of lines. The only possibility to influence
> count.fields seems via skip, but this I could only use to skip to the
> last line, which reads the file nevertheless, and I also don't know
> the number of lines in the file.
>
> Perhaps one could catch an error, when the first invocation of
> read.table fails, and try the second one. However tryCatch doesn't
> seem to make it simple to write something like
>
> E = try(expr1 otherwise expr2)
>
> (if expr1 fails, evaluate expr2 instead) ?
>
> Oliver
>
> ______________________________________________
> 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.
>



-- 
Joshua Wiley
Ph.D. Student, Health Psychology
Programmer Analyst II, ATS Statistical Consulting Group
University of California, Los Angeles
https://joshuawiley.com/