Beta Distribution#
There are two shape parameters a,b>0 and the support is x∈[0,1].
f(x;a,b)=Γ(a+b)Γ(a)Γ(b)xa−1(1−x)b−1F(x;a,b)=∫x0f(y;a,b)dy=I(x;a,b)G(q;a,b)=I−1(q;a,b)M(t)=Γ(a)Γ(b)Γ(a+b)1F1(a;a+b;t)μ=aa+bμ2=ab(a+b+1)(a+b)2γ1=2b−aa+b+2√a+b+1abγ2=6(a3+a2(1−2b)+b2(b+1)−2ab(b+2))ab(a+b+2)(a+b+3)md=(a−1)(a+b−2)a+b≠2
where I(x;a,b) is the regularized incomplete Beta function. f(x;a,1) is also called the Power-function distribution.
lx(a,b)=−NlogΓ(a+b)+NlogΓ(a)+NlogΓ(b)−N(a−1)¯logx−N(b−1)¯log(1−x)
Implementation: scipy.stats.beta