Sorry the above reply was to tutufan, not you. I deleted the quoted text 
and didn't realise it made it hard to tell who I was talking to. I agree 
feel safer to have a < check rather than ==

On Monday, May 7, 2012 10:42:12 PM UTC+1, A wrote:
>
> I see your point in general, but it is not a problem in this case for two 
> reasons:
>
> 1. the calculation is exactly the same so you are assured that you will 
> reach the exact same IEEE752 float
> 2. the termination of the while rely on an inequality "<" not on the an 
> equality "==" so even in a world with unstable
>  IEEE752 operations the loop stops with enough precision to pass the 
> Online Judge.
>
> I also just posted my solution and analysis for Equal Sums:
>
>
> http://codejamdaemon.blogspot.it/2012/05/equal-sums-google-codejam-2012-round-1b.html
>
> Which is almost identical to yours (and to what I actually used in the 
> round).
>
> A.
>
> On Monday, 7 May 2012 23:25:35 UTC+2, Chris Knott wrote:
>>
>> No, I don't think so because it relies on the <= operator working 
>> properly - which it does. Basically, you are at some point going to get the 
>> same set of numbers again in 'small', so this calculation will produce 
>> exactly the same bits, as it is the same calculation, not just "the same 
>> number" in a logical sense.
>>
>> You get problems with floats when you are trying to argue that two things 
>> are "the same number", like sqrt(16) and 4 - this sort of thing might have 
>> float problems.
>>
>

-- 
You received this message because you are subscribed to the Google Groups 
"Google Code Jam" group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-code/-/BixGYfEfjyMJ.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/google-code?hl=en.

Reply via email to