neurostuff/PyMARE: 0.0.11

Résumé fourni par la source

<!-- Release notes generated using configuration in .github/release.yml at master --> PyMARE is being given a lift, this release offers much more stability, correctness, and efficiency. with ch2, satterwaite_dof, and kh corrections being added, stan being updated, testing harness refactored, some longstanding bugs being fixed, likelihood estimators being improved by a factor of over 20x. What's Changed [ENH] model dependency between observations by @jdkent in https://github.com/neurostuff/PyMARE/pull/133 Modernize testing by @jdkent in https://github.com/neurostuff/PyMARE/pull/134 [FIX] compare the robumeta reference values numerically, not byte for… by @jdkent in https://github.com/neurostuff/PyMARE/pull/138 [PERF] Vectorize the likelihood estimators across parallel datasets by @jdkent in https://github.com/neurostuff/PyMARE/pull/137 [REF] refactor/update STAN by @jdkent in https://github.com/neurostuff/PyMARE/pull/136 [ENH] k h correction by @jdkent in https://github.com/neurostuff/PyMARE/pull/139 [FIX] handle underflow by using logp by @jdkent in https://github.com/neurostuff/PyMARE/pull/140 [FIX] concordant z is the statistic, not the inverse of its corrected p by @jdkent in https://github.com/neurostuff/PyMARE/pull/141 Full Changelog: https://github.com/neurostuff/PyMARE/compare/0.0.10...0.0.11