Fix for complex powers of symbolics

[karlwessel]

This fixes issue #717

[karlwessel]

Anything preventing this from merging?

[github-actions]

Benchmark Results

	master	835fb0d...	master / 835fb0d...
overhead/acrule/a+2	1.2 ± 0.33 μs	1.22 ± 0.35 μs	0.983 ± 0.39
overhead/acrule/a+2+b	1.18 ± 0.32 μs	1.19 ± 0.3 μs	0.991 ± 0.37
overhead/acrule/a+b	0.264 ± 0.026 μs	0.264 ± 0.11 μs	1 ± 0.43
overhead/acrule/noop:Int	2.48 ± 0.001 ns	2.47 ± 0.91 ns	1 ± 0.37
overhead/acrule/noop:Sym	27.2 ± 6 ns	0.041 ± 0.00077 μs	0.663 ± 0.15
overhead/rule/noop:Int	0.0356 ± 0.025 μs	0.0576 ± 0.026 μs	0.617 ± 0.52
overhead/rule/noop:Sym	0.0343 ± 0.025 μs	0.0687 ± 0.063 μs	0.499 ± 0.59
overhead/rule/noop:Term	0.0414 ± 0.026 μs	0.111 ± 0.062 μs	0.372 ± 0.31
overhead/ruleset/noop:Int	0.113 ± 0.0035 μs	0.118 ± 0.014 μs	0.958 ± 0.11
overhead/ruleset/noop:Sym	0.132 ± 0.0036 μs	0.156 ± 0.0076 μs	0.846 ± 0.047
overhead/ruleset/noop:Term	4.5 ± 0.51 μs	7.11 ± 1.2 μs	0.633 ± 0.13
overhead/simplify/noop:Int	0.238 ± 0.019 μs	0.228 ± 0.034 μs	1.04 ± 0.18
overhead/simplify/noop:Sym	0.235 ± 0.03 μs	0.223 ± 0.02 μs	1.06 ± 0.17
overhead/simplify/noop:Term	0.0482 ± 0.0038 ms	0.0389 ± 0.0053 ms	1.24 ± 0.19
overhead/simplify/randterm (+, *):serial	0.121 ± 0.0037 s	0.0761 ± 0.0088 s	1.59 ± 0.19
overhead/simplify/randterm (+, *):thread	0.0727 ± 0.0062 s	0.043 ± 0.0038 s	1.69 ± 0.21
overhead/simplify/randterm (/, *):serial	0.229 ± 0.022 ms	0.208 ± 0.046 ms	1.1 ± 0.27
overhead/simplify/randterm (/, *):thread	0.269 ± 0.021 ms	0.225 ± 0.049 ms	1.19 ± 0.28
overhead/substitute/a	0.106 ± 0.012 ms	0.138 ± 0.015 ms	0.766 ± 0.12
overhead/substitute/a,b	0.0909 ± 0.011 ms	0.124 ± 0.0068 ms	0.731 ± 0.096
overhead/substitute/a,b,c	21.8 ± 2.7 μs	0.068 ± 0.0046 ms	0.321 ± 0.045
polyform/easy_iszero	0.0407 ± 0.0044 ms	0.0498 ± 0.0081 ms	0.818 ± 0.16
polyform/isone	3.1 ± 0.01 ns	3.1 ± 0.01 ns	0.997 ± 0.0045
polyform/iszero	1.43 ± 0.037 ms	1.44 ± 0.044 ms	0.998 ± 0.04
polyform/simplify_fractions	1.95 ± 0.051 ms	2.12 ± 0.073 ms	0.921 ± 0.04
time_to_load	1.15 ± 0.0094 s	1.15 ± 0.011 s	1 ± 0.013

Benchmark Plots

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).

[karlwessel]

Failing integration test seems unrelated.

And I am not sure why some of the benchmarks are much slower with this PR, can't imagine that this change should have much of an effect.

[AayushSabharwal]

It's weird but also very reproducible and in some cases very significant. Could you rebase to make sure it isn't due to some outdated changes?

[karlwessel]

Sorry, something went wrong here. I rebased locally and then force pushed the changes. Somehow that closed the pull request automatically...