Re: [GiNaC-devel] Fixing log evaluation for powers

29 Dec 2013

      Hi Vladimir,

On 11/04/2013 01:57 PM, Vladimir V. Kisil wrote:
...
Making check on current master, I noticed that the following
  test
if (!ex(log(pow(p,a))).is_equal(a*log(p)))
     ++result;
from exam_inifnc.cpp failed. I think that the evaluation rule
log(p^a) -> a*log(p), if p>0 and a is real
falls out of inifcns_trans.cpp. I am attaching the corresponding
  patch.
Out of curiosity, I let git bisect to the regression and found the
culprit. But how exactly your expansion rules for exp(), log(), and
abs() managed to break this evaluation is a mystery to me.

Anyway, thanks for your patch! I've applied it.

Cheers
   -richy.
-- 
Richard B. Kreckel