On 2/27/07, didier deshommes <[EMAIL PROTECTED]> wrote:
> I must have somehow miscompiled the library (again). I will look at it
> more closely later.
There was a subtle bug in the quad double constructor that caused some
loss of precision.
This is better:
{{{
cos
mpfr vs qd 9.9466072136833891880E-64
mpfr vs rd 2.28816965375e-13
sin
mpfr vs qd 1.0255213165420582647E-63
mpfr vs rd 4.47974990436e-13
tan
mpfr vs qd 9.4955677457597987475E-63
mpfr vs rd 7.72715225139e-13
acos
mpfr vs qd 9.5145588812513183450E-63
mpfr vs rd 1.59383617415e-12
asin
mpfr vs qd 9.5667845038529972381E-63
mpfr vs rd 1.59361412955e-12
atan
mpfr vs qd 1.0961446016511467679E-63
mpfr vs rd 4.22106793962e-13
cosh
mpfr vs qd 1.2002397630640385617E-62
mpfr vs rd 2.975397706e-13
sinh
mpfr vs qd 1.2287264663013179579E-62
mpfr vs rd 5.15643083787e-13
tanh
mpfr vs qd 6.2338402250913078777E-63
mpfr vs rd 4.19553281006e-13
cosh
mpfr vs qd 1.7554018010883820896E-316
mpfr vs rd nan
sinh
mpfr vs qd 4.3584655953037476251E-63
mpfr vs rd 4.50750547998e-13
tanh
mpfr vs qd 3.2474841690498511716E-63
mpfr vs rd 1.08952846745e-11
}}}
Quad double seems to be faster on all operations now (including all
special functions and +,-,*,/). Here is an hg bundle that fixes that
and a few other bugs:
http://sage.math.washington.edu/home/dfdeshom/custom/patches/qd-20070301.hg
didier
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/
-~----------~----~----~----~------~----~------~--~---
