Open jolars opened 2 years ago
This is a good idea @jolars. Would you be interested in sending a PR with the numba modifications to the python benchmark file (perf.py)? I'd also like to see how it fairs.
I would add numba versions in addition to the plain python versions as both are of interest. Most Python uses sos not use numba, after all.
Since numba (https://numba.pydata.org/) and Julia have similar aims, it would be nice to either have the python benchmark converted to use numba altogether (I guess this is similar to the choice of using LuaJIT) or a separate numba benchmark.