fast accurate Kneser sexp algorithm

[nuninho1980]

Strangely, the "superf"/"isuperf" subroutines sort of work for bases<eta, For example, if you type
init(sqrt(2));
superf(0) You'll get 5.76705325376430. This is the superfunction developed from the upper repelling fixed point, which is "4". For bases<eta, the sexp(z) function doesn't work, nor does the loop(n) function. It would require special unique code to get the function to work for the lower fixed point (=2), since "2" is an attracting fixed point, and the rest of the code is for repelling fixed points. Maybe someday? Also, eta doesn't work at all, and the loop(n) routine overflows for bases<1.7. I haven't debugged to understand why it ges an "exponent (expo) overflow."

ya! I have same error when I tried base=1.7. I try base=sqrt(2) and more bases...

your code already was updated today. wow.