From: John Maddock (john_at_[hidden])
Date: 2006-04-24 12:51:27
> Sure. The advantage to have a library solution is that it can handle
> corner cases better than a naive implementation.
Yep, this is something lots of people write believing it to be trivial and
then end up fixing all the corner cases.
>> So I guess if you wanted a really super-duper handle it all version
>> you would make it policy based:
> I agree that we should provide as much customization as possible, but
> not more. In particular, I would use policy classes only if necessary,
> preferring multiple functors instead of one single functor with
> multiple policies. The task is perceived by a lot of programmer as
> trivial (although it isn't) and if the class is too complicated
> people might get discouraged and won't use it.
>> * Values above a threshold are regarded as effectively infinity.
> Is this necessary? I mean, that might be useful if we want to really
> compute a meaningful estimate of the relative error, but in this case
> we simply want to check if the relative error is smaller than a
> (hopefully small) value.
Actually probably not :-)
>> * Absolute errors are used if the values are small enough.
> That might be a different algorithm. I would keep both relative-error
> and relative-error-or-absolute-error-if-values-small-enough. They both
> sounds right in different use cases.
Yep, and in fact once you have the basics sorted out it's trivial to create
a thin wrapper that implements a policy like that if you need it.
>> Which of these you actually want to use in your application depend
>> on the QOI of the component you're testing I guess.
>> Personally I treat all denorms as zeros because if the result is a
>> denorm there probably isn't enough information content throughout
>> the calculation that led to the denorm to get an accurate value (if
>> you see what I mean).
> I think I understand and I partially agree. However it's not the fact
> that denormals are probably already inaccurate that worries me (the
> library should not make assumptions on how the data has been obtained)
> but rather that the division by such a small number might produce an
> inaccurate, and therefore useless, result.
Yep, good point.
BTW I notice in other messages that folks have been talking about a function
object for this, but.... what use would that be for? At present I can't see
anything simpler and easier to use than just a free function:
template <class T>
T relative_error(T, T);
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk