Closed rocky closed 3 years ago
@TiagoCavalcante @mmatera
I rebased this off of master and ran a benchmark test. Here is what I see:
$ mathics-bench -v benchmarks/bench.yaml quickpatterntest
Mathics git repo /src/external-vcs/github/Mathics3/mathics-benchmark/Mathics at bab67e
500 iterations of NextPrime...
0.069489 secs for: NextPrime[1,1]
500 iterations of ElementData...
0.138089 secs for: ElementData["Gold", "AtomicNumber"]
Dumping information to file /src/external-vcs/github/Mathics3/mathics-benchmark/mathics_benchmark/../results/bench.json
Mathics git repo /src/external-vcs/github/Mathics3/mathics-benchmark/Mathics at cb2a5b
500 iterations of NextPrime...
0.069610 secs for: NextPrime[1,1]
500 iterations of ElementData...
0.138699 secs for: ElementData["Gold", "AtomicNumber"]
I am okay with merging this in if that's what people want, but I think we are not devoting our effort to where it matters with stuff like this and would ask we focus on bigger gains.
The differences are small here too.
Do whatever you want.
Pattern-matching signature changes for NextPrime and ElementData
See https://github.com/mathics/Mathics/pull/1567 for details.