Mpmath – Python library for arbitrary-precision floating-point arithmetic

Re:Mpmath – Python library for arbitrary-precision floating-... by HerrMonnezza on Tuesday February 16, 2016 @10:32AM (#11110460)

My understanding is that pure Python implementations of "compute-intensive" algorithms are interesting now because they can run on PyPy (whereas Python modules that require C libraries can, in general, not).

[ Parent | on HN ]

Re:Mpmath – Python library for arbitrary-precision floating-... by fdej on Tuesday February 16, 2016 @12:03PM (#11111268)

Sadly, PyPy doesn't speed up mpmath a whole lot.

Edit: I just tested multiplying two 100 by 100 matrices, and got a 5x speedup with PyPy compared to CPython. Which is nice but not amazing.

[ Parent | on HN ]

Re:Mpmath – Python library for arbitrary-precision floating-... by detaro on Tuesday February 16, 2016 @07:03PM (#11114113)

I'd be curious to see how the results look if you repeat the entire thing 100 or 1000 times. PyPy takes time to warm up.

[ Parent | on HN ]

Re:Mpmath – Python library for arbitrary-precision floating-... by fdej on Tuesday February 16, 2016 @07:15PM (#11114156)

It was warmed up.

[ Parent | on HN ]

Re:Mpmath – Python library for arbitrary-precision floating-... by detaro on Tuesday February 16, 2016 @08:32PM (#11114521)

Thanks!

[ Parent | on HN ]