sandialabs / quends / 15865102001

29%

push

github

Merge pull request #91 from sandialabs/average_ensemble

This pull request introduces several important enhancements and resolves key open issues in the quends framework related to analysis traceability, result transparency, and user workflow flexibility.

Issues Addressed
#78: Return effective sample size and window size as part of statistics
All relevant analysis/statistics functions (such as steady-state detection, windowed statistics, and uncertainty estimation) now return not only the computed values, but also the effective sample size and window size used. This makes downstream interpretation of statistical summaries more reliable and reproducible, providing users with richer metadata for validation, reporting, or further analysis.

#77: Allow start_time option in all trim functions
The start_time parameter is now supported and properly propagated across all trimming-related functions. This enables advanced use cases, such as beginning trimming or steady-state detection from an arbitrary time or index, which is particularly useful for workflows that require exclusion of known transients or experimental setup intervals.

#74: Pass analysis options back along with analysis results
All analytical outputs (e.g., from trimming, statistics, or steady-state detection methods) now include a metadata block containing the exact options and parameters used for computation. This makes every output fully self-describing, enhancing reproducibility and facilitating automated audit trails or downstream meta-analyses.

Additional Improvements
Comprehensive Test Coverage:
The PR adds and updates extensive unit tests to ensure that all new behaviors are robustly validated—including edge cases (e.g., missing/invalid columns, non-stationary data, empty input, and parameter propagation).

Enhanced Output Structure:
Functions that return analysis results have been updated to return dictionaries or objects that bundle results with relevant metadata, making them easier ... (continued)

149 of 604 branches covered (24.67%)

Branch coverage included in aggregate %.

445 of 1311 relevant lines covered (33.94%)

0.34 hits per line