[R] Why does sin(pi) not return 0?

[(Ted Harding)]

On 26-Sep-2013 07:55:38 Rolf Turner wrote:
> On 09/26/13 19:31, Babak Bastan wrote:
>> Hi experts,
>>
>> If I test sin(pi) in r, it returns me 1.224606e-16
>>
>> Why doesn't return me 0?
> 
> If you think that 1.224606e-16 is different from 0, you should probably not
> be using computers.

Is that a Fortune? And, if so, should R be using computers?

  sin(pi)
  # [1] 1.224606e-16
  sin(pi)==0
  # [1] FALSE

> See FAQ 7.31 (which is in a way about the inverse of
> your question, but it should provide the necessary insight).
> 
>      cheers,
>      Rolf Turner

Though, mind you, FAQ 3.71 does also offer some consolation to R:

  all.equal(0,sin(pi))
  # [1] TRUE

So it depends on what you mean by "different from". Computers
have their own fuzzy concept of this ... Babak has too fussy
a concept.

Ted.

-------------------------------------------------
E-Mail: (Ted Harding) <Ted.Harding at wlandres.net>
Date: 26-Sep-2013  Time: 09:13:33
This message was sent by XFMail