Re: [Rd] AS_NUMERIC and as.numeric - Could someone explain?

From: Douglas Bates <bates_at_stat.wisc.edu>
Date: Fri 02 Jul 2004 - 04:27:11 EST

I knew I should have checked before I mailed that response. :-) I guess you are right and I was wrong.

I don't have the opportunity to check bits of R code on the computer that I use for email this month. (I am in northern Canada and my only email connection is via a 24 Kb modem connection on a machine running Windows ME. The modem connection is metered so I don't want to download the installer for R-1.9.1 for Windows just to be able to check a few bits of code.) I'll try to stay quiet unless I'm sure of the answer.

Simon Urbanek wrote:
> On Jul 1, 2004, at 6:09 AM, Douglas Bates wrote:
>

>> AS_NUMERIC in C more closely corresponds to as.double() in R
>>
>> All that as.numeric() in R does is assure that the result is of a 
>> numeric mode.  It does not coerce integer variables to doubles.

>
>
> Um... are you sure?? For the sake of lots of code I saw I hope you're
> wrong:
>
>> Check, for example,
>> str(as.numeric(1:100))

>
>
> Yes:
>
> > str(as.double(1:10))
> num [1:10] 1 2 3 4 5 6 7 8 9 10
> > str(as.numeric(1:10))
> num [1:10] 1 2 3 4 5 6 7 8 9 10
> > str(1:10)
> int [1:10] 1 2 3 4 5 6 7 8 9 10
>
> And moreover:
>
> > as.numeric
> function (x, ...)
> UseMethod("as.double")
> <environment: namespace:base>
>
> ... admittedly I didn't answer the original question, but I suspect that
> the problem lies somewhere else ...
>
> Simon


R-devel@stat.math.ethz.ch mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-devel Received on Fri Jul 02 04:28:21 2004

This archive was generated by hypermail 2.1.8 : Wed 03 Nov 2004 - 22:45:01 EST