Message-ID: <2248a612-6c0c-aaf6-1699-a93029b31c2d@auckland.ac.nz>
Date: 2018-11-07T20:33:37Z
From: Rolf Turner
Subject: Problem with the matrix() function.
In the course of writing a bit of somewhat convoluted code I recently
made a silly error that revealed the following phenomenon:
m <- matrix(1:10,nrow=2,ncol=c(5,4))
produces
> [,1] [,2] [,3] [,4] [,5]
> [1,] 1 3 5 7 9
> [2,] 2 4 6 8 10
That is, the nonsense value of c(5,4) for the "ncol" argument is
accepted, without comment --- the first entry of the given ncol argument
is used.
It might be argued that this is a reasonable accommodation of the user's
ineptitude. I am of the opinion that an error should be thrown if the
value of ncol is not an integer scalar.
I have also discerned that if ncol is not an integer, it is replaced by
its floor value.
Is this a Good Thing?
What do others think?
cheers,
Rolf Turner
--
Technical Editor ANZJS
Department of Statistics
University of Auckland
Phone: +64-9-373-7599 ext. 88276