Hi One of my packages needs a look-up table of pre-calculated numbers in the data directory. I would like to have the matrix as large as possible. What is the largest size matrix that would be an acceptable datafile in an R package? [ The table is a square, upper triangular matrix consisting of logs of Stirling numbers calculated by Maple. As discussed on the List a few days ago (thanks again David!) Stirling numbers are computationally challenging; one needs exact integer arithmetic for very large integers for their calculation (a 100x100 table requires ~10^157) ] -- Robin Hankin Uncertainty Analyst National Oceanography Centre, Southampton European Way, Southampton SO14 3ZH, UK tel 023-8059-7743
Robin Hankin <r.hankin at noc.soton.ac.uk> writes:> One of my packages needs a look-up table of pre-calculated > numbers in the data directory. > > What is the largest size matrix that would be an acceptable datafile in > an R package? >Do you need the entire lookup table in memory at once? If not, you could use RSQLite and put the data there. Then you would be limited by file size rather than RAM. + seth
Gabor Grothendieck
2006-Jul-26 14:41 UTC
[Rd] largest acceptable lookup table in a package
A frame of reference based on:
http://cran.r-project.org/src/contrib
is the quantiles of the size distribution of the compressed packages:
0% 5% 25% 50% 75% 95% 100%
58 5800 17000 62500 249500 1500000 15000000
On 7/26/06, Robin Hankin <r.hankin at noc.soton.ac.uk>
wrote:> Hi
>
>
> One of my packages needs a look-up table of pre-calculated
> numbers in the data directory.
>
> I would like to have the matrix as large as possible.
>
> What is the largest size matrix that would be an acceptable datafile in
> an R package?
>
>
>
> [
> The table is a square, upper triangular matrix
> consisting of logs of Stirling numbers calculated by Maple.
>
> As discussed on the List a few days ago (thanks again David!)
> Stirling numbers are computationally challenging;
> one needs exact integer arithmetic for very large
> integers for their calculation (a 100x100 table requires ~10^157)
> ]
>
>
>
>
> --
> Robin Hankin
> Uncertainty Analyst
> National Oceanography Centre, Southampton
> European Way, Southampton SO14 3ZH, UK
> tel 023-8059-7743
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
>