The DO Loop
Statistical programming in SAS with an emphasis on SAS/IML programs
A previous article introduces the skew-normal distribution and shows how to use SAS to compute the skew-normal PDF, CDF, and random variates. The present article shows how to use the Cornish-Fisher expansion to quickly compute the quantiles of the skew-normal distribution. The method is very general: it can be used
The normal distribution is symmetric, which means that it has zero skewness. In certain applications, it is useful to embed the normal distribution into a family that can model positively and negatively skewed data. Azzalini (1985) proposed a method to add a skewness parameter, α, to any symmetric family. He
The Cornish-Fisher (C-F) expansion approximates the quantiles of an arbitrary continuous univariate distribution by using only the mean, variance, skewness, and excess kurtosis of the distribution. Intuitively, you can model the shape of many common probability distributions by using these four moment-based quantities. This article shows how to use SAS