Re: [boost] [MathStatisticalDistributions]Hypergeometricdistribution

From: John Maddock (john_at_[hidden])
Date: 2008-05-06 12:53:40

Johan Råde wrote:
> Paul A Bristow wrote:
>
>> What accuracy does Johan (and others) *need* for his(their)
>> application?
>
>
> I'm working with biological data.
> Almost all the input data have errors around 10 - 30% (or more) :-)
> Two correct decimals is more than my customers need.
> But other users do of course have other needs.

Can't remember where I saw it now, but I believe folks recomend at least 6 
decimal digits precision in the stat software's calculation.  I believe the 
problem with using lgamma is you can end up with no digits correct in some 
cases :-(

John.