> The above question still stands, but otherwise, I overlooked the most
> obvious solution:
>
> dim_1 <- function(x) dim(x)
>
> Unit: nanoseconds
> expr min lq mean median uq max neval cld
> dim(x) 0 1 72.941 1 1 2696 1000 a
> base::dim(x) 11549 13474 15105.950 14245 15399 60824 1000 c
> dim_1(x) 1 771 2801.544 771 1156 1806225 1000 a
> dim_R(x) 5390 6930 8077.753 7315 8085 249069 1000 b
> dim_R_memoized(x) 1156 1926 2520.119 2310 2695 73528 1000 a
> dim_R_memoized_2(x) 385 771 1089.243 771 1156 20019 1000 a
> dim_illegal(x) 0 1 161.480 1 386 2311 1000 a
> sum(x) 10395 15784 16459.454 15785 16169 114333 1000 c
>
> So, my best shot on the original problem would now be to either use:
>
> dim2 <- function(x) dim(x)
> foo <- function(x, dim=dim2(x)) { dim }
>
> or simply avoid the name clash via:
>
> foo <- function(x, dimx=dim(x)) { dimx }
I think you'll find that baseenv()$dim(x) and .BaseNamespaceEnv$dim(x)
are about 25 times faster than base::dim(x). This doesn't seem like
it should be necessary...
Radford
______________________________________________
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
