I have a question about acceptance rejection method and really appreciate your advice:

Suppose we want to generate random variable that has probability density function $f(x)$, since we're using acceptance-rejection method, we need another probability density function $g(x)$ and constant $M$ such that $f(x)/g(x)<=M$.

Our first step is: generate random variable $y$ from $g(y)$ and a random variable $v$ from standard uniform distribution $[0,1]$

Here is my doubt: is there one to one mapping between generated random variable $y$ and $v$? In other words, are they independent OR for each $y$, it is derived by cumulative distribution function $G^{-1}(v)$, be aware that we use $v$ in the following step

Our second step is: if $v<={f(x)}/{(M*g(X))}$, accept $x=y$


They are independent.

The point is that $y$ is derived from your easily sampled distribution $g$ randomly. Now you have a random test (via $v$) that decides whether to accept $y$ or not as part of the random sample of the harder to sample $f$.

The procedure uses $M$ in the accept-reject method and whilst you can derive conservative estimates with $M$ quite high the number of rejected samples will be very high and so sampling will take a long time. Otherwise you can do some prior analysis to determine a supposed optimal underlying $g$ and low value of $M$ that will still generate a random sample with the distribution of $f$ but the number of rejected samples will be minimised.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.