1
$\begingroup$

How I can simulate a sample from log skew t distribution with lambda equals to 0?

I didn't find any packages can do it and I'm not so in order to write a function. Can someone help me to understand how to simulate a sample from that distribution?

Improve this question
$\endgroup$
3
  • 1
    $\begingroup$ You can use the rst function in the sn package. $\endgroup$ – Stéphane Laurent Dec 10 '19 at 14:29
  • $\begingroup$ that one is a skew t distribution, I'd like to have a log skew t distribution. At the moment I'm using rst function. Thank you so much for the comment! :) $\endgroup$ – TheAvenger Dec 10 '19 at 14:30
  • 1
    $\begingroup$ The log-skew is not the exponential of the skew ? (as the log-normal is the exponential of the normal). $\endgroup$ – Stéphane Laurent Dec 10 '19 at 14:35
1
$\begingroup$

If $U$ is uniform between 0 and 1, and if $F$ is a cumulative distribution function, then the random variable $X = F^{-1}(U)$ has cumulative distribution function $F$. This result is known as the inverse probability integral transformation. Therefore, to generate a value from your distribution, it suffices to draw $u$ from $U(0, 1)$ and to make the inverse transformation $F^{−1}(u)$.

PS: I do not know how tractable is the cdf of the log skew t distribution.

Improve this answer
$\endgroup$

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.