Already captured the target samples through NUTS, (need some advice) How can i find the probability of failure?

[aaronjg]

@H.Nik

Your technique is actually pretty good and i like it. But based on that, i cannot plot the components of X like what you are seeing in my figure.

If you want the bivariate plot you could generate it in the generated quantities, by generating a new normal variate. You can calculate the conditional mean and variance, but everything should work out nicely since it’s all normal.

[H.Nik]

@aaronjg

If you want the bivariate plot you could generate it in the generated quantities, by generating a new normal variate. You can calculate the conditional mean and variance, but everything should work out nicely since it’s all normal.

Thank you very much for your help, Aaron.
My original question remain unanswered which is OK…

Regards,
Hamed

[H.Nik]

@aaronjg

If you want the bivariate plot you could generate it in the generated quantities, by generating a new normal variate. You can calculate the conditional mean and variance, but everything should work out nicely since it’s all normal.

Beside that, I have a general question, are the final samples are from my posterior regarding my original code or even your code?
Posterior ~ likelihood * prior which is in our case Posterior ~ [ target += normal_lpdf(g | 0,0.5) ] * [ x ~ multi_normal(mu, Sigma)] , Is it the correct statement?

Thank you very much for your time and consideration, Aaron. I just want to double check about that.

Hamed