anemoi-datasets 0.5.24__tar.gz → 0.5.25__tar.gz

{anemoi_datasets-0.5.24 → anemoi_datasets-0.5.25}/.pre-commit-config.yaml

@@ -41,7 +41,7 @@ repos:
     - --profile black
     - --project anemoi
 - repo: https://github.com/astral-sh/ruff-pre-commit
-  rev: v0.11.4
+  rev: v0.11.12
   hooks:
   - id: ruff
     args:
@@ -65,7 +65,7 @@ repos:
   - id: docconvert
     args: ["numpy"]
 - repo: https://github.com/tox-dev/pyproject-fmt
-  rev: "v2.5.1"
+  rev: "v2.6.0"
   hooks:
   - id: pyproject-fmt
 -   repo: https://github.com/jshwi/docsig # Check docstrings against function sig

{anemoi_datasets-0.5.24 → anemoi_datasets-0.5.25}/.release-please-config.json

@@ -10,11 +10,16 @@
   "draft-pull-request": true,
   "pull-request-title-pattern": "chore${scope}: Release${component} ${version}",
   "pull-request-header": ":robot: Automated Release PR\n\nThis PR was created by `release-please` to prepare the next release. Once merged:\n\n1. A new version tag will be created\n2. A GitHub release will be published\n3. The changelog will be updated\n\nChanges to be included in the next release:",
-  "pull-request-footer": "> [!IMPORTANT]\n> :warning: Merging this PR will:\n> - Create a new release\n> - Trigger deployment pipelines\n> - Update package versions\n\n **Before merging:**\n - Ensure all tests pass\n - Review the changelog carefully\n - Get required approvals\n\n [Release-please documentation](https://github.com/googleapis/release-please)",
+  "pull-request-footer": "> [!IMPORTANT]\n> Please do not change the PR title, manifest file, or any other automatically generated content in this PR unless you understand the implications. Changes here can break the release process.\n> :warning: Merging this PR will:\n> - Create a new release\n> - Trigger deployment pipelines\n> - Update package versions\n\n **Before merging:**\n - Ensure all tests pass\n - Review the changelog carefully\n - Get required approvals\n\n [Release-please documentation](https://github.com/googleapis/release-please)",
   "packages": {
     ".": {
         "package-name": "anemoi-datasets"
     }
   },
+  "plugins": [
+    {
+      "type": "sentence-case"
+    }
+  ],
   "$schema": "https://raw.githubusercontent.com/googleapis/release-please/main/schemas/config.json"
 }

anemoi_datasets-0.5.25/.release-please-manifest.json

@@ -0,0 +1,3 @@
+{
+    ".": "0.5.25"
+}

{anemoi_datasets-0.5.24 → anemoi_datasets-0.5.25}/CHANGELOG.md

@@ -8,6 +8,20 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 Please add your functional changes to the appropriate section in the PR.
 Keep it human-readable, your future self will thank you!
 
+## [0.5.25](https://github.com/ecmwf/anemoi-datasets/compare/0.5.24...0.5.25) (2025-06-11)
+
+
+### Features
+
+* Integrating non-regular datasets in anemoi for observations. ([#306](https://github.com/ecmwf/anemoi-datasets/issues/306)) ([95a0fe4](https://github.com/ecmwf/anemoi-datasets/commit/95a0fe4bb10dc48469c0be0efad94f4d5e2a9fe8))
+
+
+### Bug Fixes
+
+* Incremental dataset build tasks called regardless of presence of debug flag in CLI code ([#294](https://github.com/ecmwf/anemoi-datasets/issues/294)) ([37afc0d](https://github.com/ecmwf/anemoi-datasets/commit/37afc0d6489f2d6c4b3ce3f9901c40e4cec5c4eb))
+* Regression in accumulations [#354](https://github.com/ecmwf/anemoi-datasets/issues/354) ([#355](https://github.com/ecmwf/anemoi-datasets/issues/355)) ([f9769d7](https://github.com/ecmwf/anemoi-datasets/commit/f9769d7944738ecbedb6b3cc1f78cd26de36a73f))
+* Remove 2 layers of build function ([#348](https://github.com/ecmwf/anemoi-datasets/issues/348)) ([7a904c4](https://github.com/ecmwf/anemoi-datasets/commit/7a904c451772089f120419a9d39bff746e0aeebb))
+
 ## [0.5.24](https://github.com/ecmwf/anemoi-datasets/compare/0.5.23...0.5.24) (2025-05-23)
 
 

{anemoi_datasets-0.5.24 → anemoi_datasets-0.5.25}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: anemoi-datasets
-Version: 0.5.24
+Version: 0.5.25
 Summary: A package to hold various functions to support training of ML models on ECMWF data.
 Author-email: "European Centre for Medium-Range Weather Forecasts (ECMWF)" <software.support@ecmwf.int>
 License:                                  Apache License

anemoi_datasets-0.5.25/docs/adr/adr-1.md

@@ -0,0 +1,63 @@
+# Support irregular observations datasets
+
+## Status
+
+<!--What is the status? -->
+
+Proposed - 30/04/2025
+
+## Context
+
+<!--What is the issue that we are seeing that is motivating this decision or change?-->
+
+The objective of this change is to support observations data which is not regular.
+
+In contrast with the fields data where each date contain the same number of points,
+in the observations data, the number of points can change for every time window.
+
+The Zarr format fits well the fields data, but does not fit the observations data.
+
+To allow storing data with irregular shape, we need to use another format than the zarr used for fields.
+An experimental implementation using xarray-zarr has been developed and is not optimised for ML training.
+
+## Decision
+
+<!--Describe the change that you are proposing.-->
+
+Add a functionality in anemoi-datasets to read observations datasets and provide the data as dictionary/mapper of numpy arrays.
+
+Mimic as much as possible what is done for field datasets :
+
+`ds = open_dataset(....)`
+`ds[i]` -> provides the data for a given time window, related to a given reference date. As a dictionary-like object.
+`ds.dates` -> list of reference date, `ds.dates[i]` is the reference date for the data provided in `ds[i]`
+
+Also expose the latitudes, longitudes in a sensible way (as `ds.latitudes` and `ds.longitudes` now depend on the dates) and name_to_index and statistics and metadata, etc.
+
+These API choices need to be made on an actual training use case.
+
+Step to achieve this:
+- Implement now a prototype format to allow developing ML training code on observation data.
+- Performing extensive benchmarking with various formats (explore parquet, and other).
+- As the final format is not defined yet, ensure a flexible architecture to allow switching (this will help for benchmarking).
+
+
+## Scope of Change
+
+<!--Specify which Anemoi packages/modules will be affected by this decision.-->
+- anemoi-datasets
+Not a breaking change, this only add functionality to read observations datasets.
+
+Must be in line with the change related to multi-datasets.
+
+## Consequences
+
+<!--Discuss the impact of this decision, including benefits, trade-offs, and potential technical debt.-->
+
+## Alternatives Considered [Optional]
+
+<!--List alternative solutions and why they were not chosen.-->
+
+## References [Optional]
+
+<!--Links to relevant discussions, documentation, or external resources.-->

{anemoi_datasets-0.5.24 → anemoi_datasets-0.5.25}/src/anemoi/datasets/_version.py

@@ -17,5 +17,5 @@ __version__: str
 __version_tuple__: VERSION_TUPLE
 version_tuple: VERSION_TUPLE
 
-__version__ = version = '0.5.24'
-__version_tuple__ = version_tuple = (0, 5, 24)
+__version__ = version = '0.5.25'
+__version_tuple__ = version_tuple = (0, 5, 25)

{anemoi_datasets-0.5.24 → anemoi_datasets-0.5.25}/src/anemoi/datasets/commands/finalise-additions.py

@@ -61,7 +61,8 @@ class FinaliseAdditions(Command):
 
         if "debug" in options:
             options.pop("debug")
-            task(step, options)
+
+        task(step, options)
 
         LOG.info(f"Create step '{step}' completed in {seconds_to_human(time.time()-now)}")
 

{anemoi_datasets-0.5.24 → anemoi_datasets-0.5.25}/src/anemoi/datasets/commands/finalise.py

@@ -55,7 +55,8 @@ class Finalise(Command):
 
         if "debug" in options:
             options.pop("debug")
-            task(step, options)
+
+        task(step, options)
 
         LOG.info(f"Create step '{step}' completed in {seconds_to_human(time.time()-now)}")
 

{anemoi_datasets-0.5.24 → anemoi_datasets-0.5.25}/src/anemoi/datasets/commands/init-additions.py

@@ -61,7 +61,8 @@ class InitAdditions(Command):
 
         if "debug" in options:
             options.pop("debug")
-            task(step, options)
+
+        task(step, options)
 
         LOG.info(f"Create step '{step}' completed in {seconds_to_human(time.time()-now)}")
 

{anemoi_datasets-0.5.24 → anemoi_datasets-0.5.25}/src/anemoi/datasets/commands/load-additions.py

@@ -62,7 +62,8 @@ class LoadAdditions(Command):
 
         if "debug" in options:
             options.pop("debug")
-            task(step, options)
+
+        task(step, options)
 
         LOG.info(f"Create step '{step}' completed in {seconds_to_human(time.time()-now)}")
 

{anemoi_datasets-0.5.24 → anemoi_datasets-0.5.25}/src/anemoi/datasets/commands/load.py

@@ -62,7 +62,8 @@ class Load(Command):
 
         if "debug" in options:
             options.pop("debug")
-            task(step, options)
+
+        task(step, options)
 
         LOG.info(f"Create step '{step}' completed in {seconds_to_human(time.time()-now)}")
 

{anemoi_datasets-0.5.24 → anemoi_datasets-0.5.25}/src/anemoi/datasets/create/__init__.py

@@ -44,7 +44,7 @@ from .check import check_data_values
 from .chunks import ChunkFilter
 from .config import build_output
 from .config import loader_config
-from .input import build_input
+from .input import InputBuilder
 from .statistics import Summary
 from .statistics import TmpStatistics
 from .statistics import check_variance
@@ -101,7 +101,9 @@ def json_tidy(o: Any) -> Any:
 
 
 def build_statistics_dates(
-    dates: list[datetime.datetime], start: Optional[datetime.datetime], end: Optional[datetime.datetime]
+    dates: list[datetime.datetime],
+    start: Optional[datetime.datetime],
+    end: Optional[datetime.datetime],
 ) -> tuple[str, str]:
     """Compute the start and end dates for the statistics.
 
@@ -551,36 +553,16 @@ class HasElementForDataMixin:
 
         self.output = build_output(config.output, parent=self)
 
-        self.input = build_input_(main_config=config, output_config=self.output)
-        # LOG.info("%s", self.input)
-
-
-def build_input_(main_config: Any, output_config: Any) -> Any:
-    """Build the input for the dataset.
-
-    Parameters
-    ----------
-    main_config : Any
-        The main configuration.
-    output_config : Any
-        The output configuration.
-
-    Returns
-    -------
-    Any
-        The input builder.
-    """
-    builder = build_input(
-        main_config.input,
-        data_sources=main_config.get("data_sources", {}),
-        order_by=output_config.order_by,
-        flatten_grid=output_config.flatten_grid,
-        remapping=build_remapping(output_config.remapping),
-        use_grib_paramid=main_config.build.use_grib_paramid,
-    )
-    LOG.debug("✅ INPUT_BUILDER")
-    LOG.debug(builder)
-    return builder
+        self.input = InputBuilder(
+            config.input,
+            data_sources=config.get("data_sources", {}),
+            order_by=self.output.order_by,
+            flatten_grid=self.output.flatten_grid,
+            remapping=build_remapping(self.output.remapping),
+            use_grib_paramid=config.build.use_grib_paramid,
+        )
+        LOG.debug("✅ INPUT_BUILDER")
+        LOG.debug(self.input)
 
 
 class Init(Actor, HasRegistryMixin, HasStatisticTempMixin, HasElementForDataMixin):
@@ -1541,7 +1523,16 @@ class Statistics(Actor, HasStatisticTempMixin, HasRegistryMixin):
         if not all(self.registry.get_flags(sync=False)):
             raise Exception(f"❗Zarr {self.path} is not fully built, not writing statistics into dataset.")
 
-        for k in ["mean", "stdev", "minimum", "maximum", "sums", "squares", "count", "has_nans"]:
+        for k in [
+            "mean",
+            "stdev",
+            "minimum",
+            "maximum",
+            "sums",
+            "squares",
+            "count",
+            "has_nans",
+        ]:
             self.dataset.add_dataset(name=k, array=stats[k], dimensions=("variable",))
 
         self.registry.add_to_history("compute_statistics_end")

{anemoi_datasets-0.5.24 → anemoi_datasets-0.5.25}/src/anemoi/datasets/create/input/__init__.py

@@ -104,23 +104,3 @@ class InputBuilder:
             Trace string.
         """
         return f"InputBuilder({group_of_dates})"
-
-
-def build_input(config: dict, data_sources: Union[dict, list], **kwargs: Any) -> InputBuilder:
-    """Build an InputBuilder instance.
-
-    Parameters
-    ----------
-    config : dict
-        Configuration dictionary.
-    data_sources : Union[dict, list]
-        Data sources.
-    **kwargs : Any
-        Additional keyword arguments.
-
-    Returns
-    -------
-    InputBuilder
-        An instance of InputBuilder.
-    """
-    return InputBuilder(config, data_sources, **kwargs)

{anemoi_datasets-0.5.24 → anemoi_datasets-0.5.25}/src/anemoi/datasets/create/sources/accumulations.py

@@ -459,12 +459,13 @@ class AccumulationFromStart(Accumulation):
             A tuple representing the MARS date-time step.
         """
         assert user_date is None, user_date
-        assert not frequency, frequency
 
         steps = (step1 + add_step, step2 + add_step)
         if steps[0] == 0:
             steps = (steps[1],)
 
+        assert frequency == 0 or frequency == (step2 - step1), frequency
+
         return (
             base_date.year * 10000 + base_date.month * 100 + base_date.day,
             base_date.hour * 100 + base_date.minute,
@@ -824,6 +825,11 @@ def _compute_accumulations(
     step1, step2 = user_accumulation_period
     assert step1 < step2, user_accumulation_period
 
+    if accumulations_reset_frequency is not None:
+        AccumulationClass = AccumulationFromLastReset
+    else:
+        AccumulationClass = AccumulationFromStart if data_accumulation_period in (0, None) else AccumulationFromLastStep
+
     if data_accumulation_period is None:
         data_accumulation_period = user_accumulation_period[1] - user_accumulation_period[0]
 
@@ -838,11 +844,6 @@ def _compute_accumulations(
 
     base_times = [t // 100 if t > 100 else t for t in base_times]
 
-    if accumulations_reset_frequency is not None:
-        AccumulationClass = AccumulationFromLastReset
-    else:
-        AccumulationClass = AccumulationFromStart if data_accumulation_period in (0, None) else AccumulationFromLastStep
-
     mars_date_time_steps = AccumulationClass.mars_date_time_steps(
         dates=dates,
         step1=step1,

{anemoi_datasets-0.5.24 → anemoi_datasets-0.5.25}/src/anemoi/datasets/data/dataset.py

@@ -179,6 +179,19 @@ class Dataset(ABC, Sized):
         if "start" in kwargs or "end" in kwargs:
             start = kwargs.pop("start", None)
             end = kwargs.pop("end", None)
+            padding = kwargs.pop("padding", None)
+
+            if padding:
+                if padding != "empty":
+                    raise ValueError(f"Only 'empty' padding is supported, got {padding=}")
+                from .padded import Padded
+
+                frequency = kwargs.pop("frequency", self.frequency)
+                return (
+                    Padded(self, start, end, frequency, dict(start=start, end=end, frequency=frequency))
+                    ._subset(**kwargs)
+                    .mutate()
+                )
 
             from .subset import Subset
 
@@ -724,6 +737,9 @@ class Dataset(ABC, Sized):
         """Return the grid shape of the dataset."""
         return (self.shape[-1],)
 
+    def empty_item(self) -> NDArray[Any]:
+        return np.zeros((*self.shape[1:-1], 0), dtype=self.dtype)
+
     def _check(self) -> None:
         """Check for overridden private methods in the dataset."""
         common = Dataset.__dict__.keys() & self.__class__.__dict__.keys()
@@ -1075,3 +1091,16 @@ class Dataset(ABC, Sized):
             The dataset names.
         """
         pass
+
+    def get_latitudes(self, i):
+        return self.get_aux(i)[0]
+
+    def get_longitudes(self, i):
+        return self.get_aux(i)[1]
+
+    def get_timedeltas(self, i):
+        return self.get_aux(i)[2]
+
+    def get_aux(self, i):
+        # need to decide if Fields datasets need to implement this
+        raise NotImplementedError(f"get_aux is not implemented for this dataset, {type(self)}")

{anemoi_datasets-0.5.24 → anemoi_datasets-0.5.25}/src/anemoi/datasets/data/misc.py

@@ -11,6 +11,7 @@
 import calendar
 import datetime
 import logging
+import os
 from pathlib import PurePath
 from typing import TYPE_CHECKING
 from typing import Any
@@ -22,6 +23,7 @@ from typing import Union
 
 import numpy as np
 import zarr
+from anemoi.utils.config import load_any_dict_format
 from anemoi.utils.config import load_config as load_settings
 from numpy.typing import NDArray
 
@@ -108,7 +110,10 @@ def round_datetime(d: np.datetime64, dates: NDArray[np.datetime64], up: bool) ->
 
 
 def _as_date(
-    d: Union[int, str, np.datetime64, datetime.date], dates: NDArray[np.datetime64], last: bool
+    d: Union[int, str, np.datetime64, datetime.date],
+    dates: NDArray[np.datetime64],
+    last: bool,
+    frequency: Optional[datetime.timedelta] = None,
 ) -> np.datetime64:
     """Convert a date to a numpy datetime64 object, rounding to the nearest date in a list of dates.
 
@@ -120,6 +125,8 @@ def _as_date(
         The list of dates.
     last : bool
         Whether to round to the last date.
+    frequency : Optional[datetime.timedelta]
+        The frequency of the dataset.
 
     Returns
     -------
@@ -142,30 +149,49 @@ def _as_date(
         pass
 
     if isinstance(d, int):
+        delta = frequency
+        if delta is None:
+            delta = np.timedelta64(1, "s")
+        delta = np.timedelta64(delta, "s")
+
         if len(str(d)) == 4:
             year = d
             if last:
-                return _as_date(np.datetime64(f"{year:04}-12-31T23:59:59"), dates, last)
+                year = year + 1
+                npdate = np.datetime64(f"{year:04}-01-01T00:00:00")
+                return _as_date(npdate - delta, dates, last, frequency)
             else:
-                return _as_date(np.datetime64(f"{year:04}-01-01T00:00:00"), dates, last)
+                return _as_date(np.datetime64(f"{year:04}-01-01T00:00:00"), dates, last, frequency)
 
         if len(str(d)) == 6:
             year = d // 100
             month = d % 100
             if last:
-                _, last_day = calendar.monthrange(year, month)
-                return _as_date(np.datetime64(f"{year:04}-{month:02}-{last_day:02}T23:59:59"), dates, last)
+                month = month + 1
+                if month > 12:
+                    month = 1
+                    year += 1
+                npdate = np.datetime64(f"{year:04}-{month:02}-01T00:00:00")
+                return _as_date(npdate - delta, dates, last, frequency)
             else:
-                return _as_date(np.datetime64(f"{year:04}-{month:02}-01T00:00:00"), dates, last)
+                return _as_date(np.datetime64(f"{year:04}-{month:02}-01T00:00:00"), dates, last, frequency)
 
         if len(str(d)) == 8:
             year = d // 10000
             month = (d % 10000) // 100
             day = d % 100
             if last:
-                return _as_date(np.datetime64(f"{year:04}-{month:02}-{day:02}T23:59:59"), dates, last)
+                day = day + 1
+                if day > calendar.monthrange(year, month)[1]:
+                    day = 1
+                    month += 1
+                    if month > 12:
+                        month = 1
+                        year += 1
+                npdate = np.datetime64(f"{year:04}-{month:02}-{day:02}T00:00:00")
+                return _as_date(npdate - delta, dates, last, frequency)
             else:
-                return _as_date(np.datetime64(f"{year:04}-{month:02}-{day:02}T00:00:00"), dates, last)
+                return _as_date(np.datetime64(f"{year:04}-{month:02}-{day:02}T00:00:00"), dates, last, frequency)
 
     if isinstance(d, str):
 
@@ -201,19 +227,20 @@ def _as_date(
                 np.datetime64(f"{year:04}-{month:02}-{day:02}T{hour:02}:{minute:02}:{second:02}"),
                 dates,
                 last,
+                frequency,
             )
 
         if "-" in d:
             assert ":" not in d
             bits = d.split("-")
             if len(bits) == 1:
-                return _as_date(int(bits[0]), dates, last)
+                return _as_date(int(bits[0]), dates, last, frequency)
 
             if len(bits) == 2:
-                return _as_date(int(bits[0]) * 100 + int(bits[1]), dates, last)
+                return _as_date(int(bits[0]) * 100 + int(bits[1]), dates, last, frequency)
 
             if len(bits) == 3:
-                return _as_date(int(bits[0]) * 10000 + int(bits[1]) * 100 + int(bits[2]), dates, last)
+                return _as_date(int(bits[0]) * 10000 + int(bits[1]) * 100 + int(bits[2]), dates, last, frequency)
 
         if ":" in d:
             assert len(d) == 5
@@ -225,12 +252,16 @@ def _as_date(
             month = first.month
             day = first.day
 
-            return _as_date(np.datetime64(f"{year:04}-{month:02}-{day:02}T{hour}:00:00"), dates, last)
+            return _as_date(np.datetime64(f"{year:04}-{month:02}-{day:02}T{hour}:00:00"), dates, last, frequency)
 
     raise NotImplementedError(f"Unsupported date: {d} ({type(d)})")
 
 
-def as_first_date(d: Union[int, str, np.datetime64, datetime.date], dates: NDArray[np.datetime64]) -> np.datetime64:
+def as_first_date(
+    d: Union[int, str, np.datetime64, datetime.date],
+    dates: NDArray[np.datetime64],
+    frequency: Optional[datetime.timedelta] = None,
+) -> np.datetime64:
     """Convert a date to the first date in a list of dates.
 
     Parameters
@@ -239,16 +270,22 @@ def as_first_date(d: Union[int, str, np.datetime64, datetime.date], dates: NDArr
         The date to convert.
     dates : NDArray[np.datetime64]
         The list of dates.
+    frequency : Optional[datetime.timedelta]
+        The frequency of the dataset.
 
     Returns
     -------
     np.datetime64
         The first date.
     """
-    return _as_date(d, dates, last=False)
+    return _as_date(d, dates, last=False, frequency=frequency)
 
 
-def as_last_date(d: Union[int, str, np.datetime64, datetime.date], dates: NDArray[np.datetime64]) -> np.datetime64:
+def as_last_date(
+    d: Union[int, str, np.datetime64, datetime.date],
+    dates: NDArray[np.datetime64],
+    frequency: Optional[datetime.timedelta] = None,
+) -> np.datetime64:
     """Convert a date to the last date in a list of dates.
 
     Parameters
@@ -257,13 +294,15 @@ def as_last_date(d: Union[int, str, np.datetime64, datetime.date], dates: NDArra
         The date to convert.
     dates : NDArray[np.datetime64]
         The list of dates.
+    frequency : Optional[datetime.timedelta]
+        The frequency of the dataset.
 
     Returns
     -------
     np.datetime64
         The last date.
     """
-    return _as_date(d, dates, last=True)
+    return _as_date(d, dates, last=True, frequency=frequency)
 
 
 def _concat_or_join(datasets: List["Dataset"], kwargs: Dict[str, Any]) -> Tuple["Dataset", Dict[str, Any]]:
@@ -317,6 +356,18 @@ def _open(a: Union[str, PurePath, Dict[str, Any], List[Any], Tuple[Any, ...]]) -
     from .stores import Zarr
     from .stores import zarr_lookup
 
+    if isinstance(a, str) and len(a.split(".")) in [2, 3]:
+
+        metadata_path = os.path.join(a, "metadata.json")
+        if os.path.exists(metadata_path):
+            metadata = load_any_dict_format(metadata_path)
+            if "backend" not in metadata:
+                raise ValueError(f"Metadata for {a} does not contain 'backend' key")
+
+            from anemoi.datasets.data.records import open_records_dataset
+
+            return open_records_dataset(a, backend=metadata["backend"])
+
     if isinstance(a, Dataset):
         return a.mutate()
 
@@ -454,6 +505,13 @@ def _open_dataset(*args: Any, **kwargs: Any) -> "Dataset":
     for a in args:
         sets.append(_open(a))
 
+    if "observations" in kwargs:
+        from .observations import observations_factory
+
+        assert not sets, sets
+
+        return observations_factory(args, kwargs).mutate()
+
     if "xy" in kwargs:
         # Experimental feature, may be removed
         from .xy import xy_factory