On this pdf paper, authors show formulas for calculating Trimmed Mean and

Winsorized Mean on unsorted data:

http://www.jstatsoft.org/v39/c02/paper

Regarding formulas (2) , (3) and (4) ... they show that formula (4) is a way

to compute formula (2) and they use min() and max() functions practically to

solve K() on formula (2) .. BUT what is K() ?

Formula (3) is Trimmed Mean and they don't give another version of it. Should

be converted to the same form of formula (4) by using min() and max() and if

yes, how should it be written correctly?

Is K() the K-Function by Ripley? And if yes, which version?

Of course there are formulas (5) and (6) where authors give another way of computing the same Trimmed Mean on unsorted data and for which they provide their C code implementation.

However I'd like to understand how K() should be solved and written on formula (3) for Trimmed Mean which uses Winsorized Mean of formula (4).

Thanks to anyone that can explain it to me.