delip(1,1)=0, instead of Inf
Reported by Roland Koberle
BUG DESCRIPTION:
----------------
The elliptic integral of first kind gives wrong result: it should diverge for k=1. As you increase k toward one, delip does in fact increase, but
eventually gives 0, instead of Inf
ERROR LOG:
----------
--> delip(1-1/10^14,1)
ans =
16.465069
--> delip(1-1/10^15,1)
ans =
17.616362
--> delip(1-1/10^16,1)
ans =
18.714974
--> delip(1-1/10^17,1)
ans =
0.
delip(1,1) should give Inf
HOW TO REPRODUCE THE BUG:
-------------------------
delip(1.0,1.0)
ans =
0.
OTHER INFORMATION:
------------------