plot synthetic skewed normal distribution

[Isabel Martinez]

Hi,
For illustration purposes I would like to plot a skew normal distribution to add to the following plot:

twoway function normalden(x, 80, 10), ra(-10 200) || function normalden(x, 80, 20), ra(-10 200)  ///
, legend(order(1 "{&sigma} = 10" 2 "{&sigma} = 20") row(1)) ///
xline(80, lcolor(grey) lpattern(dot))

Is there a way to add a skewness parameter to the density function in order to allow for non-zero skewness? Or a handy workaround?

Any suggestions are much appreciated!

[Nick Cox]

What you see in the help is what there is. Functions for the normal distribution in Stata don't include the ability to mimic anything else.

SJ-10-4 st0207 . Suite of commands for fitting skew-normal and skew-t models
. . . . . . . . . . . . . . . . . . . Y. V. Marchenko and M. G. Genton
(help skewnreg, skewtreg, mskewnreg, mskewtreg,
skew_postestimation, skewrplot if installed)
Q4/10 SJ 10(4):507--539
provides univariate and multivariate skew-normal and skew-t
regressions

may help here.

[Scott Merryman]

"A simple family of infinitely differentiable transformations that can achieve any skewness coefficient (beginning with a standard Normal distribution) is x→sgn(λ)x+exp(λx), λ∈R. "

From:
https://stats.stackexchange.com/ques...t304025_159640

[Nick Cox]

http://azzalini.stat.unipd.it/SN/ may help.