On Mon, 5 Nov 2007, Ignacio Diaz-Emparanza wrote:
I needed a "string" parameter in a function...
We're working on that; see Jack's response and my follow-up.
There is also another command that worked in the previous
version but it is not working now:
matrix A = { NA, NA, NA, NA }
Sorry, that was never supposed to work: it was a definite bug that
you could do that and not get an error message. The bug is now
fixed. The point is that we're supposed to able to do arbitrary C
floating-point arithmetic with matrices, and you can't do that
with NAs. So matrices are not allowed to contain NA. I'm pretty
sure it's in the manual.
Perhaps I should ask: why do you want NAs in a matrix? Maybe we
can think of an alternative approach.
Allin.