This function calculates the logit of proportions.
logit(p)A numeric Splus object, all of whose values are in the range [0,1]. Missing
values (NAs) are allowed.
A numeric object of the same type as p containing the logits of the input
values.
If any elements of p are outside the unit interval then an error message
is generated. Values of p equal to 0 or 1 (to within machine precision)
will return -Inf or Inf respectively. Any NAs in the input will also
be NAs in the output.