[Gretl-users] Treatment of nan by the quantile function

How does the quantile function treats/handles nan values in a series (and/or matrix column if the treatment is different)? I am asking motivated by the following. Consider <hansl> scalar bootquant95 = quantile(bootmom3Stat,0.95) scalar bootquant90 = quantile(bootmom3Stat,0.90) scalar bootquant85 = quantile(bootmom3Stat,0.85) scalar bootquant80 = quantile(bootmom3Stat,0.80) matrix bootRejmom3Stat[i,1] = (mom3Stat[i,1] > bootquant95) matrix bootRejmom3Stat[i,2] = (mom3Stat[i,1] > bootquant90) matrix bootRejmom3Stat[i,3] = (mom3Stat[i,1] > bootquant85) matrix bootRejmom3Stat[i,4] = (mom3Stat[i,1] > bootquant80) </hansl> "bootmom3Stat" (whose quantiles I have to compute) is a column vector. Some of its values are nan. The value "mom3Stat[i,1]" is always a number. What happens sometimes is that as regards the four elements of the i-row of matrix "bootRejmom3Stat", some values result as nan while others as "0" or "1". For example I have one case where (mom3Stat[i,1] > bootquant95) = nan (mom3Stat[i,1] > bootquant90) = 0 (mom3Stat[i,1] > bootquant85) = nan (mom3Stat[i,1] > bootquant80) = nan In another example, I got (mom3Stat[i,1] > bootquant95) = 1 (mom3Stat[i,1] > bootquant90) = 1 (mom3Stat[i,1] > bootquant85) = nan (mom3Stat[i,1] > bootquant80) = 0 It appears strange to have a proper number value for the 0.8-quantile and for the 0.9-quantile so that the inequality condition can be checked, but not for the 0.85-quantile. -- Alecos Papadopoulos PhD Athens University of Economics and Business web: alecospapadopoulos.wordpress.com/ scholar:https://g.co/kgs/BqH2YU