expression not in canonical form

[zimmermann6]

In some cases GiNaC leaves numeric objects in the terms of an add:

sage: a=(sqrt(4*(sqrt(3) - 5)*(sqrt(3) + 5) + 48) + 4*sqrt(3))/ (sqrt(3) + 5)
sage: a.imag()
sin(1/2*arctan2(0, -88* + 1))*sqrt(abs(4*(sqrt(3) - 5)*(sqrt(3) + 5) + 1))

See the output of the second argument of arctan2.

The term -88* is generated by the expair -22*4. This term should be moved to the overall coefficient.

See also #9913.

Depends on #11317

Upstream: Fixed upstream, in a later stable release.

Component: symbolics

Keywords: pynac

Author: Richard Kreckel, Burcin Erocal

Reviewer: Karl-Dieter Crisman

Merged: sage-4.7.1.alpha1

Issue created by migration from https://trac.sagemath.org/ticket/9947

[burcin]

add doctest

[burcin]

Author: Richard Kreckel, Burcin Erocal

[burcin]

Upstream: Fixed upstream, in a later stable release.

[burcin]

Changed keywords from none to pynac

[burcin]

comment:1

Attachment: trac_9947-add_eval.patch.gz

This was fixed in GiNaC by Richard Kreckel. Here is the relevant patch:

http://www.ginac.de/ginac.git?p=ginac.git;a=commitdiff;h=e08cda1854bdb82f6706ec269233577690ae00e4

I applied the patch to pynac, so this will be fixed in the next release.

[burcin]

comment:2

New pynac package with the fix is at #11317.

[burcin]

Dependencies: 11317

[kcrisman]

Reviewer: Karl-Dieter Crisman

[kcrisman]

comment:3

Looks ok. Same comment as at #9891.

For instance, one could then allow an spkg maintainer to review the upstream fix. But that's not exactly what we want.

[jdemeyer]

Changed dependencies from 11317 to #11317

[jdemeyer]

Merged: sage-4.7.1.alpha1