estimation parameter mixture

[a.kordba...@gmail.com]

hello everybody

i want estimation parameters mixture GB2 distribution to covariat:

i have two subgroups(k=1,2)

For a mixture of two GB2 to covariat, the likelihood can be written as:

f(y)=p1*f1(y)+(1-p1)*f2(y)

where 0<p1<1

and y[i] ~ GB2(a[k],bk[i],p[k],q[k])

that: bk[i]=E(y[i])*B(p[k],q[k])/B(p[k]+1/a[k],q[k]-1/a[k])

and: log(E(y[i]))=mu[k]+alpha[k]*F[i]+beta1[k]*x1[i]+beta2[k]*x2[i]+beta3[k]*x3[i]

Priors

mu[k]~N(0,100)

alpha[k]~N(0,100)

betaj[k]~N(0,100)

p[k]~U(0,1)

alpha[k]~N(0,100)

p[k]~Ga(.001,.001)

q[k]~Ga(.01,.0001)

best regards,