MétaCan
Menu
Back to cohort
Record W6930940673 · doi:10.5281/zenodo.3990008

Ouranosinc/xclim: v0.19.0

2020· other· en· W6930940673 on OpenAlex

Why this work is in the frame

A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.

affAt least one author lists a Canadian institution in the pinned OpenAlex snapshot.

Bibliographic record

VenueZenodo (CERN European Organization for Nuclear Research) · 2020
Typeother
Languageen
FieldBiochemistry, Genetics and Molecular Biology
TopicMachine Learning in Bioinformatics
Canadian institutionsOuranos
Fundersnot available
KeywordsCode refactoringClass (philosophy)MetadataObject (grammar)Key (lock)Variable (mathematics)Order (exchange)Set (abstract data type)Relation (database)Dimension (graph theory)

Abstract

fetched live from OpenAlex

Breaking changes Refactoring of the <code>Indicator</code> class. The <code>cfprobe</code> method has been renamed to <code>cfcheck</code> and the <code>validate</code> method has been renamed to <code>datacheck</code>. More importantly, instantiating <code>Indicator</code> creates a new subclass on the fly and stores it in a registry, allowing users to subclass existing indicators easily. The algorithm for missing values is identified by its registered name, e.g. "any", "pct", etc, along with its <code>missing_options</code>. xclim now requires xarray &gt;= 0.16, ensuring that xclim.sdba is fully functional. The dev requirements now include <code>xdoctest</code> -- a rewrite of the standard library module, <code>doctest</code>. <code>xclim.core.locales.get_local_attrs</code> now uses the indicator's class name instead of the indicator itself and no longer accepts the <code>fill_missing</code> keyword. Behaviour is now the same as passing <code>False</code>. <code>Indicator.cf_attrs</code> is now a list of dictionaries. <code>Indicator.json</code> puts all the metadata attributes in the key "outputs" (a list of dicts). All variable metadata (names in <code>Indicator._cf_names</code>) might be strings or lists of strings when accessed as object attributes. Passing doctests are now strictly enforced as a build requirement in the Travis CI testing ensemble. New features and enhancements New <code>ensembles.kkz_reduce_ensemble</code> method to select subsets of an ensemble based on the KKZ algorithm. Create new Indicator <code>Daily</code>, <code>Daily2D</code> subclasses for indicators using daily input data. The <code>Indicator</code> class now supports outputing multiple indices for the same inputs. <code>xclim.core.units.declare_units</code> now works with indices outputting multiple DataArrays. Doctests now make use of the <code>xdoctest_namespace</code> in order to more easily access mdoules and tesdata. Bug fixes Fix <code>generic.fit</code> dimension ordering. This caused errors when "time" was not the first dimension in a DataArray. Internal changes <code>datachecks.check_daily</code> now uses <code>xr.infer_freq</code>. Indicator subclasses <code>Tas</code>, <code>Tasmin</code>, <code>Tasmax</code>, <code>Pr</code> and <code>Streamflow</code> now inherit from <code>Daily</code>. Indicator subclasses <code>TasminTasmax</code> and <code>PrTas</code> now inherit from <code>Daily2D</code>. Docstring style now enforced using the <code>pydocstyle</code> with <code>numpy</code> doctsring conventions. Doctests are now performed for all docstring <code>Examples</code> using <code>xdoctest</code>. Failing examples must be explicitly skipped otherwise build will now fail. Indicator methods <code>update_attrs</code> and <code>format</code> are now classmethods, attrs to update must be passed. Indicators definitions without an accompanying translation (presently French) will cause build failures. Major refactoring of the internal marchinery of <code>Indicator</code> to support multiple outputs.

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.

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.000
metaresearch head score (Gemma)0.001
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesMeta-epidemiology (narrow), Insufficient payload (model declined to judge)
Consensus categoriesInsufficient payload (model declined to judge)
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Not applicable · Consensus signal: Not applicable
GenreCandidate signal: Other · Consensus signal: Other
Teacher disagreement score0.092
Threshold uncertainty score1.000

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0000.001
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0000.000
Science and technology studies0.0010.000
Scholarly communication0.0000.000
Open science0.0010.001
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0230.011

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.

Opus teacher head0.019
GPT teacher head0.248
Teacher spread0.229 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it