4 of 6 standards met
Summary: Remove pyre-fixme/pyre-ignore suppression comments from 11 files across several directories: plot/, analysis/, generation_strategy/, benchmark/, metrics/, global_stopping/ (both source and test files). Key fixes: Use for Optional unwrapping Use for type narrowing Add proper type annotations (, , ) Use wrapping for numpy scalar arithmetic Use instead of bare Use for cardinality() return values Fix -> for RGB color tuples Annotate global Remove stale pyre-fixme comments on calls Differential Revision: D95264987
Repository: facebook/Ax. Description: Adaptive Experimentation Platform Stars: 2712, Forks: 361. Primary language: Python. Languages: Python (95.2%), Jupyter Notebook (4.2%), JavaScript (0.3%), Makefile (0.1%), Batchfile (0.1%). License: MIT. Homepage: https://ax.dev Latest release: 1.2.4 (21h ago). Open PRs: 100, open issues: 79. Last activity: 4m ago. Community health: 75%. Top contributors: saitcakmak, mpolson64, lena-kashtelyan, esantorella, bernardbeckerman, sdaulton, mgarrard, ldworkin, Balandat, 2timesjay and others.
Last 12 weeks · 266 commits
Summary: crashes with when has zero rows (e.g., during initial Sobol trial generation before any trials have completed). This happens because on an empty DataFrame returns an empty DataFrame rather than an empty Series. Assigning that DataFrame to a single column then fails. Replace with a list comprehension over , which correctly produces an empty list for empty DataFrames. Failing run: https://www.internalfb.com/mlhub/flow/1044918711/overview Differential Revision: D95266027
Summary: The exception message is now displayed in the ErrorAnalysisCard subtitle, allowing users to immediately see the reasoning (e.g., why an analysis is not applicable) without needing to expand the error card. Previously, the subtitle only showed a generic message like "AnalysisNotApplicableStateError encountered while computing ParallelCoordinatesPlot." Now it shows the actual exception message, such as "AnalysisNotApplicableStateError: Experiment has no trials", providing immediate context. Differential Revision: D94617700
Summary: Analysis card to show transferrable experiments with a default of 25% parameter overlap. Differential Revision: D92926519