casevh
casevh
> I'm wondering if this approach could work for python-flint. We would still need to build libflint though and I am not sure if that could be done with MSVC...
I fixed many performance issues with mixed operands in what will be released as 2.1.0b6 (or possibly as 2.1.0rc1). If you are running on Linux and can compile gmpy2 from...
I reconstructed the code from your first message. The latest gmpy2 code isn't appreciably faster. I'm now curious...
I got it closer but mpmath still wins. The MPFR library implements an optimized (unsigned long) ** (unsigned long) function that I was not using. I added that special case...
Nice work. Once you reach "large enough" precision it usually doesn't make much difference whether you implement an algorithm in Python or C. The MPFR library uses a similar approach...
I have a 12-core server with some idle time. :)
I have done some performance tests with the new build process. One of the goals of `gmpy2` has been to reduce overhead to a minimum. I used the test suite...
That appears to be the cause. I didn't expect the import overhead to be that high. I will write a more specific test over the weekend.
The behavior you are seeing is the desired behavior -- by default, mpfr() and mpc() should convert exactly if possible.This changes the rules around conversion but should reduce errors due...
I'm planning to make the following change in 2.1.0.b4: 1) If possible, `mpfr` and `mpc` conversions will be done exactly. 2) If the precision of the result is less than...