Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

CompatHelper: bump compat for Turing in [weakdeps] to 0.36, (keep existing compat) #111

Conversation

github-actions[bot]
Copy link
Contributor

This pull request changes the compat entry for the Turing package from 0.29.0,0.30.0,0.31.0,0.32,0.33,0.34.0,0.35.0 to 0.29.0,0.30.0,0.31.0,0.32,0.33,0.34.0,0.35.0, 0.36.
This keeps the compat entries for earlier versions.

Note: I have not tested your package with this new compat entry.
It is your responsibility to make sure that your package tests pass before you merge this pull request.

@itsdfish itsdfish closed this Jan 16, 2025
@itsdfish itsdfish reopened this Jan 16, 2025
Copy link
Contributor Author

Benchmark Results

master 9fb93ee... master/9fb93ee091240c...
logpdf/("SequentialSamplingModels.DDM", 10) 1.68 ± 0.18 μs 1.69 ± 0.18 μs 0.998
logpdf/("SequentialSamplingModels.DDM", 100) 17.3 ± 0.99 μs 17.2 ± 0.78 μs 1
logpdf/("SequentialSamplingModels.LBA", 10) 2.46 ± 0.21 μs 2.46 ± 0.2 μs 0.999
logpdf/("SequentialSamplingModels.LBA", 100) 24 ± 0.69 μs 23.9 ± 0.69 μs 1
logpdf/("SequentialSamplingModels.LNR", 10) 0.969 ± 0.17 μs 0.974 ± 0.17 μs 0.994
logpdf/("SequentialSamplingModels.LNR", 100) 8.46 ± 0.31 μs 8.5 ± 0.32 μs 0.996
logpdf/("SequentialSamplingModels.RDM", 10) 2.64 ± 0.25 μs 2.64 ± 0.24 μs 1
logpdf/("SequentialSamplingModels.RDM", 100) 25.3 ± 0.84 μs 25.3 ± 0.81 μs 1
logpdf/("SequentialSamplingModels.Wald", 10) 0.235 ± 0.18 μs 0.238 ± 0.17 μs 0.987
logpdf/("SequentialSamplingModels.Wald", 100) 2.05 ± 0.072 μs 2.09 ± 0.055 μs 0.981
logpdf/("SequentialSamplingModels.WaldMixture", 10) 1.12 ± 0.19 μs 1.12 ± 0.17 μs 1
logpdf/("SequentialSamplingModels.WaldMixture", 100) 10.8 ± 0.18 μs 10.8 ± 0.18 μs 1
rand/("SequentialSamplingModels.DDM", 10) 2.91 ± 0.38 μs 2.89 ± 0.38 μs 1.01
rand/("SequentialSamplingModels.DDM", 100) 28 ± 1.3 μs 28 ± 1.3 μs 0.999
rand/("SequentialSamplingModels.LBA", 10) 3.09 ± 1.4 μs 3.06 ± 1.2 μs 1.01
rand/("SequentialSamplingModels.LBA", 100) 14.9 ± 0.67 μs 14.9 ± 0.85 μs 1
rand/("SequentialSamplingModels.LCA", 10) 0.593 ± 0.19 ms 0.603 ± 0.18 ms 0.983
rand/("SequentialSamplingModels.LCA", 100) 6.4 ± 0.23 ms 6.52 ± 0.25 ms 0.982
rand/("SequentialSamplingModels.LNR", 10) 1.03 ± 0.16 μs 1.02 ± 0.17 μs 1.01
rand/("SequentialSamplingModels.LNR", 100) 7.66 ± 3.9 μs 7.55 ± 3.8 μs 1.01
rand/("SequentialSamplingModels.RDM", 10) 1.37 ± 0.16 μs 1.37 ± 0.18 μs 0.999
rand/("SequentialSamplingModels.RDM", 100) 11.3 ± 3.9 μs 11.2 ± 3.9 μs 1.01
rand/("SequentialSamplingModels.Wald", 10) 0.471 ± 0.17 μs 0.466 ± 0.17 μs 1.01
rand/("SequentialSamplingModels.Wald", 100) 2.93 ± 0.23 μs 2.91 ± 0.22 μs 1.01
rand/("SequentialSamplingModels.WaldMixture", 10) 1.25 ± 0.17 μs 1.25 ± 0.16 μs 0.999
rand/("SequentialSamplingModels.WaldMixture", 100) 11.8 ± 0.22 μs 11.8 ± 0.22 μs 1
simulate/SequentialSamplingModels.DDM 3.88 ± 1.6 μs 3.93 ± 1.6 μs 0.986
simulate/SequentialSamplingModels.LBA 4.6 ± 4.4 μs 4.64 ± 4.5 μs 0.992
simulate/SequentialSamplingModels.LCA 0.0807 ± 0.015 ms 0.0797 ± 0.016 ms 1.01
simulate/SequentialSamplingModels.RDM 0.0749 ± 0.024 ms 0.0775 ± 0.024 ms 0.965
simulate/SequentialSamplingModels.Wald 8.91 ± 3.9 μs 9.02 ± 3.6 μs 0.987
simulate/SequentialSamplingModels.WaldMixture 4.07 ± 1.7 μs 4.07 ± 1.6 μs 1
simulate/mdft 0.179 ± 0.066 ms 0.176 ± 0.066 ms 1.02
time_to_load 1.51 ± 0.018 s 1.52 ± 0.018 s 0.997

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).

@itsdfish itsdfish closed this Jan 16, 2025
@itsdfish itsdfish deleted the compathelper/new_version/2025-01-16-01-29-34-515-01950373248 branch January 16, 2025 08:58
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant