More Realistic Benchmark

[ewanwm]

Now randomise the osc parameters at each step in the benchmark rather than the energies. This more closely matches the real life usage pattern for an oscillation calculator.

This seems to be a lot slower than previously when testing locally...

[github-actions]

Cpp-linter Review

Used clang-tidy v14.0.0

Have any feedback or feature suggestions? Share it here.

[github-actions]

clang-tidy diagnostic

benchmarks/benchmarks.cpp:11:27: warning: [readability-identifier-length]

variable name 'i' is too short, expected at least 3 characters

const std::complex<float> i(0.0, 1.0);
                          ^

[github-actions]

clang-tidy diagnostic

benchmarks/benchmarks.cpp:24:62: warning: [bugprone-easily-swappable-parameters]

2 adjacent parameters of 'build' of similar type ('const Tensor &') are easily swapped by mistake

    void build(const Tensor &theta12, const Tensor &theta13, const Tensor &theta23, const Tensor &deltaCP)
                                                             ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/home/runner/work/nuTens/nuTens/benchmarks/benchmarks.cpp:24:76: note: the first parameter in the range is 'theta23'
    void build(const Tensor &theta12, const Tensor &theta13, const Tensor &theta23, const Tensor &deltaCP)
                                                                           ^~~~~~~
/home/runner/work/nuTens/nuTens/benchmarks/benchmarks.cpp:24:99: note: the last parameter in the range is 'deltaCP'
    void build(const Tensor &theta12, const Tensor &theta13, const Tensor &theta23, const Tensor &deltaCP)
                                                                                                  ^~~~~~~

[github-actions]

clang-tidy diagnostic

benchmarks/benchmarks.cpp:56:12: warning: [cppcoreguidelines-non-private-member-variables-in-classes]

member variable 'matrix' has public visibility

    Tensor matrix;
           ^

[github-actions]

clang-tidy diagnostic

benchmarks/benchmarks.cpp:65:86: warning: [bugprone-easily-swappable-parameters]

2 adjacent parameters of 'batchedOscProbs' of similar type ('long') are easily swapped by mistake

                            Tensor &deltaCP, Tensor &masses, const Tensor &energies, long batchSize, long nBatches)
                                                                                     ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/home/runner/work/nuTens/nuTens/benchmarks/benchmarks.cpp:65:91: note: the first parameter in the range is 'batchSize'
                            Tensor &deltaCP, Tensor &masses, const Tensor &energies, long batchSize, long nBatches)
                                                                                          ^~~~~~~~~
/home/runner/work/nuTens/nuTens/benchmarks/benchmarks.cpp:65:107: note: the last parameter in the range is 'nBatches'
                            Tensor &deltaCP, Tensor &masses, const Tensor &energies, long batchSize, long nBatches)
                                                                                                          ^~~~~~~~
/home/runner/work/nuTens/nuTens/benchmarks/benchmarks.cpp:79:64: warning: 2.0 is a magic number; consider replacing it with a named constant [cppcoreguidelines-avoid-magic-numbers,readability-magic-numbers]
        deltaCP.setValue({0}, Tensor::scale(Tensor::rand({1}), 2.0 * 3.1415));
                                                               ^
/home/runner/work/nuTens/nuTens/benchmarks/benchmarks.cpp:79:70: warning: 3.1415 is a magic number; consider replacing it with a named constant [cppcoreguidelines-avoid-magic-numbers,readability-magic-numbers]
        deltaCP.setValue({0}, Tensor::scale(Tensor::rand({1}), 2.0 * 3.1415));
                                                                     ^
/home/runner/work/nuTens/nuTens/benchmarks/benchmarks.cpp:97:102: warning: 10000.0 is a magic number; consider replacing it with a named constant [cppcoreguidelines-avoid-magic-numbers,readability-magic-numbers]
        Tensor::scale(Tensor::rand({state.range(0), 1}).dType(NTdtypes::kFloat).requiresGrad(false), 10000.0) +
                                                                                                     ^
/home/runner/work/nuTens/nuTens/benchmarks/benchmarks.cpp:111:30: warning: 295000.0 is a magic number; consider replacing it with a named constant [cppcoreguidelines-avoid-magic-numbers,readability-magic-numbers]
    Propagator vacuumProp(3, 295000.0);
                             ^
/home/runner/work/nuTens/nuTens/benchmarks/benchmarks.cpp:130:102: warning: 10000.0 is a magic number; consider replacing it with a named constant [cppcoreguidelines-avoid-magic-numbers,readability-magic-numbers]
        Tensor::scale(Tensor::rand({state.range(0), 1}).dType(NTdtypes::kFloat).requiresGrad(false), 10000.0) +
                                                                                                     ^
/home/runner/work/nuTens/nuTens/benchmarks/benchmarks.cpp:145:30: warning: 295000.0 is a magic number; consider replacing it with a named constant [cppcoreguidelines-avoid-magic-numbers,readability-magic-numbers]
    Propagator matterProp(3, 295000.0);
                             ^
/home/runner/work/nuTens/nuTens/benchmarks/benchmarks.cpp:149:100: warning: 2.6 is a magic number; consider replacing it with a named constant [cppcoreguidelines-avoid-magic-numbers,readability-magic-numbers]
    std::shared_ptr<BaseMatterSolver> matterSolver = std::make_shared<ConstDensityMatterSolver>(3, 2.6);
                                                                                                   ^

[codecov-commenter]

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

All modified and coverable lines are covered by tests ✅

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.