These are chat archives for symengine/symengine

27th
Jan 2016
Isuru Fernando
@isuruf
Jan 27 2016 06:09
@certik, can you post the url of the symengine.rb appveyor badge?
Ralf Stephan
@rwst
Jan 27 2016 16:01
@isuruf Thanks!
@isuruf Um, you did mean the superfluous move?
Isuru Fernando
@isuruf
Jan 27 2016 16:06
No, the coeff
Add::from_dict(one, std::move(dict_)) one here
Ralf Stephan
@rwst
Jan 27 2016 16:08
why, it is one
ah that's not a factor
Akash Trehan
@CodeMaxx
Jan 27 2016 16:34
@isuruf can you review #736
Isuru Fernando
@isuruf
Jan 27 2016 16:38
@CodeMaxx, for #736, you really should implement it for all integers p. Not just for primes. Reason is that a user might first check if a is a quadratic residue first and then calculate it later. But quadratic residue is already calculated in the is a quadratic residue step
Akash Trehan
@CodeMaxx
Jan 27 2016 16:40
@isuruf It works for all integers. For primes it uses the legendre functions for others it uses nthmod_root
Isuru Fernando
@isuruf
Jan 27 2016 16:41
Yes, it works. I was saying that you should implement for non-primes without the usage of nthmod_root
Akash Trehan
@CodeMaxx
Jan 27 2016 16:54
@isuruf I'll take a look at this.
During my research on this topic I had landed on this link . (See the Answer 2 by poncho) So I thought not much could be done as improvement for non-primes as as had been done for the primes.
Isuru Fernando
@isuruf
Jan 27 2016 16:56
Of course we need to factorize p, which is not so easy for very large numbers, but just use the prime_factor_multiplicities function to get the factors.