Closed bowenszhu closed 1 month ago
master | fc7509587c2550... | master/fc7509587c2550... | |
---|---|---|---|
overhead/acrule/a+2 | 0.768 ± 0.016 μs | 0.726 ± 0.02 μs | 1.06 |
overhead/acrule/a+2+b | 0.735 ± 0.013 μs | 0.703 ± 0.019 μs | 1.04 |
overhead/acrule/a+b | 0.254 ± 0.0089 μs | 0.253 ± 0.012 μs | 1 |
overhead/acrule/noop:Int | 25.9 ± 0.05 ns | 25.3 ± 0.05 ns | 1.02 |
overhead/acrule/noop:Sym | 0.0366 ± 0.0056 μs | 0.0365 ± 0.0053 μs | 1 |
overhead/rule/noop:Int | 0.0438 ± 0.0011 μs | 0.0451 ± 0.0011 μs | 0.97 |
overhead/rule/noop:Sym | 0.055 ± 0.0033 μs | 0.0588 ± 0.0023 μs | 0.936 |
overhead/rule/noop:Term | 0.0547 ± 0.003 μs | 0.0585 ± 0.0021 μs | 0.935 |
overhead/ruleset/noop:Int | 0.126 ± 0.0032 μs | 0.127 ± 0.0031 μs | 0.991 |
overhead/ruleset/noop:Sym | 0.147 ± 0.0044 μs | 0.149 ± 0.0053 μs | 0.987 |
overhead/ruleset/noop:Term | 3.02 ± 0.1 μs | 3.06 ± 0.11 μs | 0.987 |
overhead/simplify/noop:Int | 0.167 ± 0.0012 μs | 0.149 ± 0.0021 μs | 1.12 |
overhead/simplify/noop:Sym | 0.184 ± 0.0077 μs | 0.17 ± 0.0041 μs | 1.08 |
overhead/simplify/noop:Term | 0.0356 ± 0.0016 ms | 0.0348 ± 0.0016 ms | 1.02 |
overhead/simplify/randterm (+, *):serial | 0.0848 ± 0.001 s | 0.0851 ± 0.00089 s | 0.997 |
overhead/simplify/randterm (+, *):thread | 0.051 ± 0.03 s | 0.0509 ± 0.029 s | 1 |
overhead/simplify/randterm (/, *):serial | 0.212 ± 0.0076 ms | 0.205 ± 0.007 ms | 1.03 |
overhead/simplify/randterm (/, *):thread | 0.239 ± 0.0091 ms | 0.234 ± 0.0084 ms | 1.02 |
overhead/substitute/a | 0.0588 ± 0.0015 ms | 0.0603 ± 0.0017 ms | 0.975 |
overhead/substitute/a,b | 0.052 ± 0.0015 ms | 0.0521 ± 0.0014 ms | 0.997 |
overhead/substitute/a,b,c | 16.7 ± 0.6 μs | 17.2 ± 0.6 μs | 0.97 |
polyform/easy_iszero | 28.6 ± 1.9 μs | 28.8 ± 2.1 μs | 0.993 |
polyform/isone | 3.1 ± 0.01 ns | 2.79 ± 0.01 ns | 1.11 |
polyform/iszero | 1.1 ± 0.029 ms | 1.09 ± 0.028 ms | 1.01 |
polyform/simplify_fractions | 1.61 ± 0.04 ms | 1.6 ± 0.04 ms | 1.01 |
time_to_load | 2.11 ± 0.0054 s | 2.12 ± 0.013 s | 0.999 |
A plot of the benchmark results have been uploaded as an artifact to the workflow run for this PR. Go to "Actions"->"Benchmark a pull request"->[the most recent run]->"Artifacts" (at the bottom).
The original test for the
where
clause in rules incorrectly handled the scope of the symbolic variablea
. Two distinct symbolic variables with the same name were created, leading to unexpected behavior when checking for object equality in the_f
function.This PR moves the definition of
_f
inside the@testset
block, ensuring that it references the same symbolic variablea
as the rule definition. This correction aligns the test with the intended behavior of thewhere
clause.