Skip to content

climate_ref.results.values #

Typed, ORM-free surface for metric-value results.

Reader is the intent-named entry point that notebooks and (eventually) the API use, via its values sub-reader. ValuesReader's read methods return frozen, detached value objects wrapped in collections that offer to_pandas(). The objects are fully materialised, so they remain valid after the originating session closes. A notebook can build a DataFrame inside a with Database(...) block and keep using it afterwards.

Everything here is a thin layer over climate_ref.results._query (the shared Select builders) and climate_ref.results.outliers. power users who need the raw Select or ORM rows use those modules directly.

Facet #

Distinct values observed for one CV dimension across a filtered query.

Source code in packages/climate-ref/src/climate_ref/results/values.py 
51
52
53
54
55
56
57
58
59
@attrs.frozen(kw_only=True)
class Facet:
    """Distinct values observed for one CV dimension across a filtered query."""

    key: str
    """The CV dimension name."""

    values: tuple[str, ...]
    """Distinct values observed for the dimension."""

key instance-attribute #

The CV dimension name.

values instance-attribute #

Distinct values observed for the dimension.

Reader #

Typed entry point to REF query results.

Constructed from a Database, which owns the session and the read-only story. This is a thin entry point: it exposes per-domain sub-readers as properties, values for metric-value reads, executions for execution-group reads, datasets for dataset reads, diagnostics for diagnostic reads, and artifacts for output path resolution (only available when a results root is supplied).

Source code in packages/climate-ref/src/climate_ref/results/values.py 
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
class Reader:
    """
    Typed entry point to REF query results.

    Constructed from a [Database][climate_ref.database.Database], which owns the session and the
    read-only story. This is a thin entry point: it exposes per-domain sub-readers as properties,
    [values][climate_ref.results.values.Reader.values] for metric-value reads,
    [executions][climate_ref.results.values.Reader.executions] for execution-group reads,
    [datasets][climate_ref.results.values.Reader.datasets] for dataset reads,
    [diagnostics][climate_ref.results.values.Reader.diagnostics] for diagnostic reads, and
    [artifacts][climate_ref.results.values.Reader.artifacts] for output path resolution
    (only available when a ``results`` root is supplied).
    """

    def __init__(self, database: Database, results: Path | None = None) -> None:
        self._db = database
        self._results = results

    @property
    def session(self) -> Session:
        """The underlying database session."""
        return self._db.session

    @functools.cached_property
    def values(self) -> ValuesReader:
        """Metric-value reads (scalar/series/facets)."""
        return ValuesReader(self._db)

    @functools.cached_property
    def executions(self) -> ExecutionsReader:
        """Execution-group and execution reads."""
        return ExecutionsReader(self._db)

    @functools.cached_property
    def datasets(self) -> "DatasetsReader":
        """Dataset reads."""
        from climate_ref.results.datasets import DatasetsReader  # noqa: PLC0415

        return DatasetsReader(self._db)

    @functools.cached_property
    def diagnostics(self) -> "DiagnosticsReader":
        """Diagnostic reads."""
        from climate_ref.results.diagnostics import DiagnosticsReader  # noqa: PLC0415

        return DiagnosticsReader(self._db)

    @functools.cached_property
    def artifacts(self) -> "ArtifactsReader":
        """
        Output path resolution.

        Raises ``ValueError`` when no ``results`` root was supplied to the constructor.
        """
        if self._results is None:
            raise ValueError(
                "reader.artifacts requires a results root; construct "
                "Reader(database, results=config.paths.results)."
            )
        from climate_ref.results.artifacts import ArtifactsReader  # noqa: PLC0415

        return ArtifactsReader(self._results)

artifacts cached property #

Output path resolution.

Raises ValueError when no results root was supplied to the constructor.

datasets cached property #

Dataset reads.

diagnostics cached property #

Diagnostic reads.

executions cached property #

Execution-group and execution reads.

session property #

The underlying database session.

values cached property #

Metric-value reads (scalar/series/facets).

ScalarValue #

A single scalar metric value, detached from the ORM.

Source code in packages/climate-ref/src/climate_ref/results/values.py 
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
@attrs.frozen(kw_only=True)
class ScalarValue:
    """A single scalar metric value, detached from the ORM."""

    id: int
    """Primary key of the underlying ``ScalarMetricValue`` row."""

    execution_id: int
    """The execution that produced this value."""

    execution_group_id: int
    """The execution group the producing execution belongs to."""

    value: float | None
    """The scalar value itself."""

    kind: str
    """Model/reference kind, resolved from the ``kind`` CV dimension (defaults to ``"model"``)."""

    dimensions: Mapping[str, str]
    """
    CV dimension values (e.g. ``source_id``, ``statistic``, ``metric``) for this row.

    ``kind`` is excluded here and surfaced as the dedicated ``kind`` field instead.
    """

    attributes: Mapping[str, Any]
    """Free-form, non-CV attributes attached to the row."""

    is_outlier: bool | None = None
    """Per-row outlier verdict; ``None`` unless outlier detection ran."""

    verification_status: str | None = None
    """``"verified"``/``"unverified"`` status paired with ``is_outlier``; ``None`` unless detection ran."""

    diagnostic_slug: str | None = None
    """Owning diagnostic's slug, populated only when ``include_context=True``."""

    provider_slug: str | None = None
    """Owning provider's slug, populated only when ``include_context=True``."""

attributes instance-attribute #

Free-form, non-CV attributes attached to the row.

diagnostic_slug = None class-attribute instance-attribute #

Owning diagnostic's slug, populated only when include_context=True.

dimensions instance-attribute #

CV dimension values (e.g. source_id, statistic, metric) for this row.

kind is excluded here and surfaced as the dedicated kind field instead.

execution_group_id instance-attribute #

The execution group the producing execution belongs to.

execution_id instance-attribute #

The execution that produced this value.

id instance-attribute #

Primary key of the underlying ScalarMetricValue row.

is_outlier = None class-attribute instance-attribute #

Per-row outlier verdict; None unless outlier detection ran.

kind instance-attribute #

Model/reference kind, resolved from the kind CV dimension (defaults to "model").

provider_slug = None class-attribute instance-attribute #

Owning provider's slug, populated only when include_context=True.

value instance-attribute #

The scalar value itself.

verification_status = None class-attribute instance-attribute #

"verified"/"unverified" status paired with is_outlier; None unless detection ran.

ScalarValueCollection #

An immutable page of scalar values plus collection-level, outlier-aware metadata.

Source code in packages/climate-ref/src/climate_ref/results/values.py 
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
@attrs.frozen(kw_only=True)
class ScalarValueCollection:
    """An immutable page of scalar values plus collection-level, outlier-aware metadata."""

    items: tuple[ScalarValue, ...]
    """The scalar values on this page."""

    total_count: int
    """Total values matching the filter before ``offset``/``limit`` (after outlier removal, if any)."""

    facets: tuple[Facet, ...]
    """Distinct CV dimension values observed over the full filtered set, before pagination."""

    offset: int
    """Rows skipped before this page."""

    limit: int | None
    """Page size requested, or ``None`` when the whole result was returned."""

    had_outliers: bool
    """Whether outlier detection ran and flagged at least one value."""

    outlier_count: int
    """Number of values flagged as outliers, or ``0`` when detection did not run."""

    def __iter__(self) -> Iterator[ScalarValue]:
        return iter(self.items)

    def __len__(self) -> int:
        return len(self.items)

    def facets_dict(self) -> dict[str, list[str]]:
        """Facets as a plain ``{dimension: [values]}`` mapping."""
        return {f.key: list(f.values) for f in self.facets}

    def to_pandas(self) -> pd.DataFrame:
        """Tidy DataFrame: one row per value, one column per CV dimension present, plus metadata."""
        detection_ran = any(v.is_outlier is not None for v in self.items)
        return scalar_values_to_frame(self.items, detection_ran=detection_ran)

facets instance-attribute #

Distinct CV dimension values observed over the full filtered set, before pagination.

had_outliers instance-attribute #

Whether outlier detection ran and flagged at least one value.

items instance-attribute #

The scalar values on this page.

limit instance-attribute #

Page size requested, or None when the whole result was returned.

offset instance-attribute #

Rows skipped before this page.

outlier_count instance-attribute #

Number of values flagged as outliers, or 0 when detection did not run.

total_count instance-attribute #

Total values matching the filter before offset/limit (after outlier removal, if any).

facets_dict() #

Facets as a plain {dimension: [values]} mapping.

Source code in packages/climate-ref/src/climate_ref/results/values.py 
185
186
187
def facets_dict(self) -> dict[str, list[str]]:
    """Facets as a plain ``{dimension: [values]}`` mapping."""
    return {f.key: list(f.values) for f in self.facets}

to_pandas() #

Tidy DataFrame: one row per value, one column per CV dimension present, plus metadata.

Source code in packages/climate-ref/src/climate_ref/results/values.py 
189
190
191
192
def to_pandas(self) -> pd.DataFrame:
    """Tidy DataFrame: one row per value, one column per CV dimension present, plus metadata."""
    detection_ran = any(v.is_outlier is not None for v in self.items)
    return scalar_values_to_frame(self.items, detection_ran=detection_ran)

SeriesValue #

A 1-d series metric value, detached from the ORM. The index is snapshotted from the shared axis.

Source code in packages/climate-ref/src/climate_ref/results/values.py 
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
@attrs.frozen(kw_only=True)
class SeriesValue:
    """A 1-d series metric value, detached from the ORM. The index is snapshotted from the shared axis."""

    id: int
    """Primary key of the underlying ``SeriesMetricValue`` row."""

    execution_id: int
    """The execution that produced this value."""

    execution_group_id: int
    """The execution group the producing execution belongs to."""

    values: Sequence[float | int]
    """The series' y-values, in index order."""

    index: Sequence[float | int | str] | None
    """The shared index axis values, snapshotted at read time."""

    index_name: str | None
    """Name of the shared index axis (e.g. a time or latitude label)."""

    reference_id: str | None
    """Set for observation/reference series; ``None`` for model series."""

    kind: str
    """Model/reference kind, resolved from the ``kind`` CV dimension (defaults to ``"model"``)."""

    dimensions: Mapping[str, str]
    """
    CV dimension values (e.g. ``source_id``, ``statistic``, ``metric``) for this row.

    ``kind`` is excluded here and surfaced as the dedicated ``kind`` field instead.
    """

    attributes: Mapping[str, Any]
    """Free-form, non-CV attributes attached to the row."""

    diagnostic_slug: str | None = None
    """Owning diagnostic's slug, populated only when ``include_context=True``."""

    provider_slug: str | None = None
    """Owning provider's slug, populated only when ``include_context=True``."""

    @property
    def is_reference(self) -> bool:
        """True for observation/reference series (``reference_id`` is set)."""
        return self.reference_id is not None

attributes instance-attribute #

Free-form, non-CV attributes attached to the row.

diagnostic_slug = None class-attribute instance-attribute #

Owning diagnostic's slug, populated only when include_context=True.

dimensions instance-attribute #

CV dimension values (e.g. source_id, statistic, metric) for this row.

kind is excluded here and surfaced as the dedicated kind field instead.

execution_group_id instance-attribute #

The execution group the producing execution belongs to.

execution_id instance-attribute #

The execution that produced this value.

id instance-attribute #

Primary key of the underlying SeriesMetricValue row.

index instance-attribute #

The shared index axis values, snapshotted at read time.

index_name instance-attribute #

Name of the shared index axis (e.g. a time or latitude label).

is_reference property #

True for observation/reference series (reference_id is set).

kind instance-attribute #

Model/reference kind, resolved from the kind CV dimension (defaults to "model").

provider_slug = None class-attribute instance-attribute #

Owning provider's slug, populated only when include_context=True.

reference_id instance-attribute #

Set for observation/reference series; None for model series.

values instance-attribute #

The series' y-values, in index order.

SeriesValueCollection #

An immutable page of series values plus collection-level metadata.

Source code in packages/climate-ref/src/climate_ref/results/values.py 
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
@attrs.frozen(kw_only=True)
class SeriesValueCollection:
    """An immutable page of series values plus collection-level metadata."""

    items: tuple[SeriesValue, ...]
    """The series values on this page."""

    total_count: int
    """Total values matching the filter before ``offset``/``limit``."""

    facets: tuple[Facet, ...]
    """Distinct CV dimension values observed over the full filtered set, before pagination."""

    offset: int
    """Rows skipped before this page."""

    limit: int | None
    """Page size requested, or ``None`` when the whole result was returned."""

    def __iter__(self) -> Iterator[SeriesValue]:
        return iter(self.items)

    def __len__(self) -> int:
        return len(self.items)

    def facets_dict(self) -> dict[str, list[str]]:
        """Facets as a plain ``{dimension: [values]}`` mapping."""
        return {f.key: list(f.values) for f in self.facets}

    def to_pandas(self, *, explode: bool = True) -> pd.DataFrame:
        """
        DataFrame of the series values.

        With ``explode=True`` (default) the result is long-form: one row per (series, index point),
        matching the API's CSV shape. With ``explode=False`` each series is one row with list-valued
        ``values``/``index`` cells.
        """
        return series_values_to_frame(self.items, explode=explode)

facets instance-attribute #

Distinct CV dimension values observed over the full filtered set, before pagination.

items instance-attribute #

The series values on this page.

limit instance-attribute #

Page size requested, or None when the whole result was returned.

offset instance-attribute #

Rows skipped before this page.

total_count instance-attribute #

Total values matching the filter before offset/limit.

facets_dict() #

Facets as a plain {dimension: [values]} mapping.

Source code in packages/climate-ref/src/climate_ref/results/values.py 
220
221
222
def facets_dict(self) -> dict[str, list[str]]:
    """Facets as a plain ``{dimension: [values]}`` mapping."""
    return {f.key: list(f.values) for f in self.facets}

to_pandas(*, explode=True) #

DataFrame of the series values.

With explode=True (default) the result is long-form: one row per (series, index point), matching the API's CSV shape. With explode=False each series is one row with list-valued values/index cells.

Source code in packages/climate-ref/src/climate_ref/results/values.py 
224
225
226
227
228
229
230
231
232
def to_pandas(self, *, explode: bool = True) -> pd.DataFrame:
    """
    DataFrame of the series values.

    With ``explode=True`` (default) the result is long-form: one row per (series, index point),
    matching the API's CSV shape. With ``explode=False`` each series is one row with list-valued
    ``values``/``index`` cells.
    """
    return series_values_to_frame(self.items, explode=explode)

ValuesReader #

Metric-value read domain: scalar/series values and their facets.

Constructed from a Database, which owns the session and the read-only story. All read methods return detached collections that outlive the session.

Source code in packages/climate-ref/src/climate_ref/results/values.py 
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
class ValuesReader:
    """
    Metric-value read domain: scalar/series values and their facets.

    Constructed from a [Database][climate_ref.database.Database], which owns the session and the
    read-only story. All read methods return detached collections that outlive the session.
    """

    def __init__(self, database: Database) -> None:
        self._db = database

    @property
    def session(self) -> Session:
        """The underlying database session."""
        return self._db.session

    def _facets(self, base_stmt: Any, entity: Any) -> tuple[Facet, ...]:
        facet_map = collect_facets(self.session, base_stmt, entity)
        return tuple(Facet(key=k, values=tuple(v)) for k, v in facet_map.items())

    def scalar_values(  # noqa: PLR0913
        self,
        filters: MetricValueFilter | None = None,
        *,
        outliers: OutlierPolicy | None = None,
        include_unverified: bool = False,
        offset: int = 0,
        limit: int | None = None,
        with_facets: bool = True,
        include_context: bool = False,
    ) -> ScalarValueCollection:
        """
        Query scalar values, returning an outlier-aware collection.

        When outlier detection is disabled (the default) pagination and counting happen in SQL,
        so cost scales with the requested page rather than the whole result set. When ``outliers``
        is enabled, detection runs over the FULL filtered set so IQR bounds are globally
        consistent; with ``include_unverified`` False, flagged values are then removed before
        pagination and excluded from ``total_count``. ``facets`` are always computed over the full
        filtered set (before any outlier removal and pagination).
        """
        filters = filters or MetricValueFilter()
        policy = outliers or OutlierPolicy(method="off")

        base_stmt = select_scalar_values(filters)
        facets = self._facets(base_stmt, ScalarMetricValue) if with_facets else ()

        if not policy.enabled:
            # No detection: page and count in SQL so we never materialise the whole table.
            total_count = count_values(self.session, base_stmt)
            load_stmt = base_stmt.options(self._scalar_loader(include_context))
            if limit is not None:
                load_stmt = load_stmt.offset(offset).limit(limit)
            elif offset:
                load_stmt = load_stmt.offset(offset)
            rows = list(self.session.execute(load_stmt).scalars().all())
            items = tuple(
                self._to_scalar_dto(r, False, "verified", include_context, detection_ran=False) for r in rows
            )
            return ScalarValueCollection(
                items=items,
                total_count=total_count,
                facets=facets,
                offset=offset,
                limit=limit,
                had_outliers=False,
                outlier_count=0,
            )

        # Detection enabled: materialise the full filtered set so IQR bounds are globally
        # consistent, then filter and paginate in Python.
        load_stmt = base_stmt.options(self._scalar_loader(include_context))
        rows = list(self.session.execute(load_stmt).scalars().all())

        annotated, outlier_count = detect_scalar_outliers(rows, policy)
        had_outliers = outlier_count > 0
        if not include_unverified:
            annotated = [a for a in annotated if not a.is_outlier]

        total_count = len(annotated)
        page = annotated[offset : offset + limit] if limit is not None else annotated[offset:]
        items = tuple(
            self._to_scalar_dto(
                a.value, a.is_outlier, a.verification_status, include_context, detection_ran=True
            )
            for a in page
        )
        return ScalarValueCollection(
            items=items,
            total_count=total_count,
            facets=facets,
            offset=offset,
            limit=limit,
            had_outliers=had_outliers,
            outlier_count=outlier_count,
        )

    def series_values(
        self,
        filters: MetricValueFilter | None = None,
        *,
        offset: int = 0,
        limit: int | None = None,
        with_facets: bool = True,
        include_context: bool = False,
    ) -> SeriesValueCollection:
        """
        Query series values with SQL-level pagination.

        The shared index axis is resolved so ``index``/``index_name`` are populated. ``facets`` are
        computed over the full filtered set before pagination.
        """
        filters = filters or MetricValueFilter()

        base_stmt = select_series_values(filters)
        facets = self._facets(base_stmt, SeriesMetricValue) if with_facets else ()
        total_count = count_values(self.session, base_stmt)

        load_stmt = base_stmt.options(self._series_loader(include_context))
        if limit is not None:
            load_stmt = load_stmt.offset(offset).limit(limit)
        elif offset:
            load_stmt = load_stmt.offset(offset)
        rows = list(self.session.execute(load_stmt).scalars().unique().all())

        items = tuple(self._to_series_dto(r, include_context) for r in rows)
        return SeriesValueCollection(
            items=items, total_count=total_count, facets=facets, offset=offset, limit=limit
        )

    # -- loaders / materialisers ------------------------------------------------
    @staticmethod
    def _scalar_loader(include_context: bool) -> Any:
        exec_load = joinedload(ScalarMetricValue.execution)
        if include_context:
            return (
                exec_load.joinedload(Execution.execution_group)
                .joinedload(ExecutionGroup.diagnostic)
                .joinedload(Diagnostic.provider)
            )
        return exec_load

    @staticmethod
    def _series_loader(include_context: bool) -> Any:
        exec_load = joinedload(SeriesMetricValue.execution)
        if include_context:
            return (
                exec_load.joinedload(Execution.execution_group)
                .joinedload(ExecutionGroup.diagnostic)
                .joinedload(Diagnostic.provider)
            )
        return exec_load

    @staticmethod
    def _context_slugs(execution: Execution, include_context: bool) -> tuple[str | None, str | None]:
        if not include_context:
            return None, None
        diagnostic = execution.execution_group.diagnostic
        return diagnostic.slug, diagnostic.provider.slug

    def _to_scalar_dto(
        self,
        row: ScalarMetricValue,
        is_outlier: bool,
        verification_status: str,
        include_context: bool,
        detection_ran: bool,
    ) -> ScalarValue:
        diagnostic_slug, provider_slug = self._context_slugs(row.execution, include_context)
        dims = dict(row.dimensions)
        kind = _kind_of(dims)
        dims.pop("kind", None)
        return ScalarValue(
            id=row.id,
            execution_id=row.execution_id,
            execution_group_id=row.execution.execution_group_id,
            value=row.value,
            kind=kind,
            dimensions=dims,
            attributes=dict(row.attributes or {}),
            is_outlier=is_outlier if detection_ran else None,
            verification_status=verification_status if detection_ran else None,
            diagnostic_slug=diagnostic_slug,
            provider_slug=provider_slug,
        )

    def _to_series_dto(self, row: SeriesMetricValue, include_context: bool) -> SeriesValue:
        diagnostic_slug, provider_slug = self._context_slugs(row.execution, include_context)
        dims = dict(row.dimensions)
        kind = _kind_of(dims)
        dims.pop("kind", None)
        return SeriesValue(
            id=row.id,
            execution_id=row.execution_id,
            execution_group_id=row.execution.execution_group_id,
            values=list(row.values or []),
            index=list(row.index) if row.index is not None else None,
            index_name=row.index_name,
            reference_id=row.reference_id,
            kind=kind,
            dimensions=dims,
            attributes=dict(row.attributes or {}),
            diagnostic_slug=diagnostic_slug,
            provider_slug=provider_slug,
        )

session property #

The underlying database session.

scalar_values(filters=None, *, outliers=None, include_unverified=False, offset=0, limit=None, with_facets=True, include_context=False) #

Query scalar values, returning an outlier-aware collection.

When outlier detection is disabled (the default) pagination and counting happen in SQL, so cost scales with the requested page rather than the whole result set. When outliers is enabled, detection runs over the FULL filtered set so IQR bounds are globally consistent; with include_unverified False, flagged values are then removed before pagination and excluded from total_count. facets are always computed over the full filtered set (before any outlier removal and pagination).

Source code in packages/climate-ref/src/climate_ref/results/values.py 
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
def scalar_values(  # noqa: PLR0913
    self,
    filters: MetricValueFilter | None = None,
    *,
    outliers: OutlierPolicy | None = None,
    include_unverified: bool = False,
    offset: int = 0,
    limit: int | None = None,
    with_facets: bool = True,
    include_context: bool = False,
) -> ScalarValueCollection:
    """
    Query scalar values, returning an outlier-aware collection.

    When outlier detection is disabled (the default) pagination and counting happen in SQL,
    so cost scales with the requested page rather than the whole result set. When ``outliers``
    is enabled, detection runs over the FULL filtered set so IQR bounds are globally
    consistent; with ``include_unverified`` False, flagged values are then removed before
    pagination and excluded from ``total_count``. ``facets`` are always computed over the full
    filtered set (before any outlier removal and pagination).
    """
    filters = filters or MetricValueFilter()
    policy = outliers or OutlierPolicy(method="off")

    base_stmt = select_scalar_values(filters)
    facets = self._facets(base_stmt, ScalarMetricValue) if with_facets else ()

    if not policy.enabled:
        # No detection: page and count in SQL so we never materialise the whole table.
        total_count = count_values(self.session, base_stmt)
        load_stmt = base_stmt.options(self._scalar_loader(include_context))
        if limit is not None:
            load_stmt = load_stmt.offset(offset).limit(limit)
        elif offset:
            load_stmt = load_stmt.offset(offset)
        rows = list(self.session.execute(load_stmt).scalars().all())
        items = tuple(
            self._to_scalar_dto(r, False, "verified", include_context, detection_ran=False) for r in rows
        )
        return ScalarValueCollection(
            items=items,
            total_count=total_count,
            facets=facets,
            offset=offset,
            limit=limit,
            had_outliers=False,
            outlier_count=0,
        )

    # Detection enabled: materialise the full filtered set so IQR bounds are globally
    # consistent, then filter and paginate in Python.
    load_stmt = base_stmt.options(self._scalar_loader(include_context))
    rows = list(self.session.execute(load_stmt).scalars().all())

    annotated, outlier_count = detect_scalar_outliers(rows, policy)
    had_outliers = outlier_count > 0
    if not include_unverified:
        annotated = [a for a in annotated if not a.is_outlier]

    total_count = len(annotated)
    page = annotated[offset : offset + limit] if limit is not None else annotated[offset:]
    items = tuple(
        self._to_scalar_dto(
            a.value, a.is_outlier, a.verification_status, include_context, detection_ran=True
        )
        for a in page
    )
    return ScalarValueCollection(
        items=items,
        total_count=total_count,
        facets=facets,
        offset=offset,
        limit=limit,
        had_outliers=had_outliers,
        outlier_count=outlier_count,
    )

series_values(filters=None, *, offset=0, limit=None, with_facets=True, include_context=False) #

Query series values with SQL-level pagination.

The shared index axis is resolved so index/index_name are populated. facets are computed over the full filtered set before pagination.

Source code in packages/climate-ref/src/climate_ref/results/values.py 
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
def series_values(
    self,
    filters: MetricValueFilter | None = None,
    *,
    offset: int = 0,
    limit: int | None = None,
    with_facets: bool = True,
    include_context: bool = False,
) -> SeriesValueCollection:
    """
    Query series values with SQL-level pagination.

    The shared index axis is resolved so ``index``/``index_name`` are populated. ``facets`` are
    computed over the full filtered set before pagination.
    """
    filters = filters or MetricValueFilter()

    base_stmt = select_series_values(filters)
    facets = self._facets(base_stmt, SeriesMetricValue) if with_facets else ()
    total_count = count_values(self.session, base_stmt)

    load_stmt = base_stmt.options(self._series_loader(include_context))
    if limit is not None:
        load_stmt = load_stmt.offset(offset).limit(limit)
    elif offset:
        load_stmt = load_stmt.offset(offset)
    rows = list(self.session.execute(load_stmt).scalars().unique().all())

    items = tuple(self._to_series_dto(r, include_context) for r in rows)
    return SeriesValueCollection(
        items=items, total_count=total_count, facets=facets, offset=offset, limit=limit
    )