Actually I didn't notice that. It seems weird to find erf in math, but erf for complex numbers in scipy.special. It's just about organization and user discovery. I realize that from the developer's point of view, erf for complex numbers is complicated, but why does the user care? On Mon, Oct 14, 2013 at 8:29 AM, Mark Dickinson <dickinsm@gmail.com> wrote:
On Sun, Oct 13, 2013 at 9:56 PM, Neil Girdhar <mistersheik@gmail.com>wrote:
By complete I meant that just as if you were to add the "error function, erf" to math, you would want to add an equivalent version to cmath.
An interesting choice of example. *Why* would you want to do so?
Since you bring this up, I assume you're already aware that math.erf exists but cmath.erf does not. I believe there are good, practical reasons *not* to add cmath.erf, in spite of the existence of math.erf. Not least of these is that cmath.erf would be significantly more complicated to implement and of significantly less interest to users. And perhaps there's a parallel with itertools.permutations and the proposed itertools.multiset_permutations here...
-- Mark