Depends what you mean by pooling. Adding uniforms gives a known
distribution - its messy but the exact distribution was (I believe)
derived by
Mitra in SIAM J Applied Math Vol 19 pp 195-198 (1970)
On 4 Aug 2009, at 10:19, Anabela Afonso wrote:
> Dear All,
>
> I'm pooling k uniform distributions with different parameters, say
> X_{i} ~Unif
> (a, b_{i}) for i = 1, ..., k.
>
> I simulate some data and grafically the simple frequencies seems
> something
> like a uniform between "a" and "min(b_{i})", and then it starts to
> fall until "max
> (b_{i})".
>
> Does anyone knows what is this theorical distibution and where can
> I find
> something about it?
>
> Thanks to all in advanced.
>
> Best wishes,
> Anabela Afonso
Gareth Janacek
01603 591206
|