quantilese

determines quantiles of data and their standard errors or confidence intervals

Usage:

[q,se]=quantilese(x,p)
[q,ql,qu]=quantilese(x,p,a)

This determines the quantiles corresponding to probability P in the data set X, along with either their standard errors or confidence intervals at level A. The standard errors are calculated using the Maritz-Jarrett method and the confidence intervals are calculated by normal approximation.

NOTE: Q1=quantilese(X,P) is not the same as Q2=prctile(X,100*P). Here Q1=XS(M1) whereas Q2 is interpolated between XS(floor(M2)) and XS(ceil(M2)) where X is of length N, XS is the sorted X, M1=round(N*P) and M2=N*P+1/2.

This function does not require the Statistics toolbox.