This routine accepts an array consisting a series of
(vectorised) lines of data. The weighted mean and standard
deviation of each input column in STACK is then used to estimate
the range of values which represent the required sigma
clipping. The standard deviation is derived from the population
of values at each position along the lines (cf. each image
pixel). Values outside of this range are then rejected and the
resulting output mean values are returned in the array RESULT.
Note that clipping will not be used when only two or three
values are available (unless in the case of 3 values NSIGMA
is less than 1.0).