> < ^ Date: Wed, 07 Sep 1994 09:56:00 -0700 (PDT)
> < ^ From: Frank Celler <frank.celler@math.rwth-aachen.de >
> < ^ Subject: Re: possible bug in square roots -- ER(N)

Dear Jacob,

This is either a bug in the function ER(N) ("square root of N")
or I might be using the function outside its defined domain :

the manual requires that N is a positive integer (see "ATALS
irrationalities"), but I agree 'ER' should either signal an error or
give the expected result or the manual should be even more definite in
restricting the domain.

However, one could write a square root function as

SquareRootInt := function(N)
    if N < 0  then
        return EI(-N);
    elif 0 < N  then
        return ER(N);
        return 0;

best wishes

> < [top]