Bibliographic record
Abstract
Contributors to this version: Abel Aoun (@bzah), Pascal Bourgault (@aulemahal), Trevor James Smith (@Zeitsperre), Gabriel Rondeau-Genesse (@RondeauG), Juliette Lavoie (@juliettelavoie), Ludwig Lierhammer (@ludwiglierhammer). Announcements xclim is now compliant with PEP 563. Python3.10-style annotations are now permitted. (GH/1065, PR/1071). xclim is now fully compatible with xarray's flox-enabled GroupBy and resample operations. (PR/1081). xclim now (properly) enforces docstring compliance checks using pydocstyle with modified numpy-style docstrings. Docstring errors will now cause build failures. See the pydocstyle documentation for more information. (PR/1074). xclim now uses GitHub Actions to manage patch version bumping. Merged Pull Requests that modify xclim code now trigger version-bumping automatically when pushed to the main development branch. Running $ bump2version patch within development branches is no longer necessary. (PR/1102). New features and enhancements Add "Celsius" to aliases of "celsius" unit. (GH/1067, PR/1068). All indicators now have indexing enabled, except those computing statistics on spells. (GH/1069, PR/1070). A convenience function for returning the version numbers for relevant xclim dependencies (xclim.testing.show_versions) is now offered. (PR/1073). A CLI version of this function is also available from the command line ($ xclim show_version_info). (PR/1073). New "keep_attrs" option to control the handling of the attributes within the indicators. (GH/1026, PR/1076). Added a notebook showcasing some simple examples of Spatial Analogues. (GH/585, PR/1075). create_ensembles now accepts a glob string to find datasets. (PR/1081). Improved percentile based indicators metadata with the window, threshold and climatology period used to compute percentiles. (GH/1047, PR/1050). New xclim.core.calendar.construct_offset, the inverse operation of parse_offset. (PR/1090). Rechunking operations in xclim.indices.run_length.rle are now synchronized with dask's options. (PR/1090). A mention of the "missing" checks and options is added to the history attribute of indicators, where appropriate. (GH/1100, PR/1103). Breaking changes xclim.atmos.water_budget has been separated into water_budget (calculated directly with 'evspsblpot') and water_budget_from_tas (original function). (PR/1086). Injected parameters in indicators are now left out of a function's signature and will not be included in the history attribute. (PR/1086). The signature for the following Indicators have been modified (PR/1050): cold_spell_duration_index, tg90p, tg10p, tx90p, tx10p, tn90p, tn10p, warm_spell_duration_index, days_over_precip_doy_thresh, days_over_precip_thresh, fraction_over_precip_doy_thresh, fraction_over_precip_thresh, cold_and_dry_days, warm_and_dry_days, warm_and_wet_days, cold_and_wet_days The parameter for percentile values is now named after the variable it is supposed to be computed upon. (PR/1050). pytest-runner has been removed as a dependency (it was never needed for xclim development). (PR/1074). xclim.testing._utils.py has been renamed to xclim.testing.utils.py for added documentation visibility. (PR/1074). Some unused functions and classes (as_tuple, TestFile, TestDataSet) have been removed. (PR/1107). New indicators universal_thermal_climate_index and mean_radiant_temperature for computing the universal thermal climate index from the near-surface temperature, relative humidity, near-surface windspeed and radiation. (GH/1060, PR/1062). A new method ITS90 has also been added for calculating saturation water vapour pressure. (GH/1060, PR/1062). Internal changes Typing syntax has been updated within pre-commit via isort. Pre-commit hooks now append from __future__ import annotations to all python module imports for backwards compatibility. (GH/1065, PR/1071) isort project configurations are now set in setup.cfg. (PR/1071). Many function docstrings, external target links, and internal section references have been adjusted to reduce warnings when building the docs. (PR/1074). Code snippets within documentation are now checked and reformatted to black conventions with blackdoc. A pre-commit hook is now in place to run these checks. (PR/1098). Test coverage statistic no longer includes coverage of the test files themselves. Coverage now reflects lines of usable code covered. (PR/1101). Reordered listed authors alphabetically. Promoted @bzah to core contributor. (PR/1105). Tests have been added for some functions in xclim.testing.utils.py; some previously uncaught bugs in list_input_variables, publish_release_notes, and show_versions have been patched. (GH/1078, PR/1107). A convenience command for installing xclim with key development branches of some dependencies has been added ($ make upstream). (GH/1088, PR/1092; amended in GH/1113, PR/1114). This build configuration is also available in tox for local development purposes ($ tox -e pyXX-upstream). Bug fixes Clean the bias_adjustement and history attributes created by xclim.sdba.adjust (e.g. when an argument is an xr.DataArray, only print the name instead of the whole array). (GH/1083, PR/1087). pydocstyle checks were silently failing in the pre-commit configuration due to a badly-formed regex. This has been adjusted. (PR/1074). adjust_doy_calendar was broken when the source or the target were seasonal. (GH/1097, GH/1091, PR/1099)
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.011 |
| Meta-epidemiology (narrow) | 0.004 | 0.003 |
| Meta-epidemiology (broad) | 0.002 | 0.003 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.005 | 0.005 |
| Open science | 0.007 | 0.008 |
| Research integrity | 0.002 | 0.005 |
| Insufficient payload (model declined to judge) | 0.428 | 0.559 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".