From: SourceForge.net <no...@so...> - 2012-11-15 03:34:19
|
Bugs item #3587184, was opened at 2012-11-14 10:19 Message generated for change (Comment added) made by rtoy You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=104933&aid=3587184&group_id=4933 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: Lisp Core - Floating point Group: None >Status: Closed >Resolution: Fixed Priority: 5 Private: No Submitted By: Raymond Toy (rtoy) Assigned to: Nobody/Anonymous (nobody) Summary: erf inaccurate for small float values Initial Comment: erf(1d-10) -> 1.128e-40 But the taylor series says erf(x) = 2*x/sqrt(%pi) so erf(1d-10) should be 1.128e-20. The error is in slatec:derf which is using the wrong formula when x < sqeps. ---------------------------------------------------------------------- >Comment By: Raymond Toy (rtoy) Date: 2012-11-14 19:34 Message: Fixed by modifying derf.f and derf.lisp ---------------------------------------------------------------------- Comment By: Raymond Toy (rtoy) Date: 2012-11-14 10:20 Message: Change summary to mention this is just for float values. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=104933&aid=3587184&group_id=4933 |