# erf(x) is the error function;
# erf(n,x) is the nth derivative of the error function;

erf(-x_):= -erf(x);