lamindb 0.76.2__py3-none-any.whl → 0.76.3__py3-none-any.whl

lamindb/__init__.py

@@ -25,7 +25,7 @@ Key functionality
 
    context
    connect
-   Curate
+   Curator
    view
    save
 
@@ -44,7 +44,7 @@ Modules & settings
 """
 
 # denote a release candidate for 0.1.0 with 0.1rc1, 0.1a1, 0.1b1, etc.
-__version__ = "0.76.2"
+__version__ = "0.76.3"
 
 import os as _os
 
@@ -94,7 +94,7 @@ if _check_instance_setup(from_lamindb=True):
         _ulabel,
         integrations,
     )
-    from ._curate import Curate
+    from ._curate import Curator
     from ._save import save
     from ._view import view
     from .core._context import context
@@ -110,6 +110,7 @@ if _check_instance_setup(from_lamindb=True):
 
     track = context.track  # backward compat
     finish = context.finish  # backward compat
+    Curate = Curator  # backward compat
     settings.__doc__ = """Global :class:`~lamindb.core.Settings`."""
     context.__doc__ = """Global :class:`~lamindb.core.Context`."""
     from django.db.models import Q

lamindb/_artifact.py

@@ -545,6 +545,7 @@ def __init__(artifact: Artifact, *args, **kwargs):
     skip_check_exists = (
         kwargs.pop("skip_check_exists") if "skip_check_exists" in kwargs else False
     )
+    _uid = kwargs.pop("_uid", None)
     if "default_storage" in kwargs:
         default_storage = kwargs.pop("default_storage")
     else:
@@ -560,6 +561,9 @@ def __init__(artifact: Artifact, *args, **kwargs):
     if "is_new_version_of" in kwargs:
         logger.warning("`is_new_version_of` will be removed soon, please use `revises`")
         revises = kwargs.pop("is_new_version_of")
+    assert not (  # noqa: S101
+        revises is not None and _uid is not None
+    ), "Can not init with both `revises` and `_uid`"
     if not len(kwargs) == 0:
         raise ValueError(
             "Only data, key, run, description, version, revises, visibility"
@@ -575,8 +579,10 @@ def __init__(artifact: Artifact, *args, **kwargs):
         raise ValueError(
             f"`key` is {key}, but `revises.key` is '{revises.key}'\n\n Either do *not* pass `key`.\n\n{note}"
         )
-
-    provisional_uid, revises = create_uid(revises=revises, version=version)
+    if _uid is not None:
+        provisional_uid, revises = _uid, None
+    else:
+        provisional_uid, revises = create_uid(revises=revises, version=version)
     if revises is not None:
         if not isinstance(revises, Artifact):
             raise TypeError("`revises` has to be of type `Artifact`")

lamindb/_curate.py

@@ -84,10 +84,34 @@ class CurateLookup:
             return colors.warning("No fields are found!")
 
 
-class DataFrameCurator:
+class BaseCurator:
+    """Curate a dataset."""
+
+    def validate(self) -> bool:
+        """Validate dataset.
+
+        Returns:
+            Boolean indicating whether the dataset is validated.
+        """
+        pass
+
+    def save_artifact(self, description: str | None = None, **kwargs) -> Artifact:
+        """Save the dataset as artifact.
+
+        Args:
+            description: Description of the DataFrame object.
+            **kwargs: Object level metadata.
+
+        Returns:
+            A saved artifact record.
+        """
+        pass
+
+
+class DataFrameCurator(BaseCurator):
     """Curation flow for a DataFrame object.
 
-    See also :class:`~lamindb.Curate`.
+    See also :class:`~lamindb.Curator`.
 
     Args:
         df: The DataFrame object to curate.
@@ -101,7 +125,7 @@ class DataFrameCurator:
 
     Examples:
         >>> import bionty as bt
-        >>> curate = ln.Curate.from_df(
+        >>> curate = ln.Curator.from_df(
         ...     df,
         ...     categoricals={
         ...         "cell_type_ontology_id": bt.CellType.ontology_id,
@@ -184,6 +208,7 @@ class DataFrameCurator:
             using_key=self._using_key,
             validated_only=False,
             source=self._sources.get("columns"),
+            exclude=self._exclude.get("columns"),
             **kwargs,
         )
 
@@ -199,6 +224,7 @@ class DataFrameCurator:
                 validated_only=validated_only,
                 df=self._df,  # Get the Feature type from df
                 source=self._sources.get("columns"),
+                exclude=self._exclude.get("columns"),
                 warning=False,  # Do not warn about missing columns, just an info message
                 **kwargs,
             )
@@ -251,6 +277,7 @@ class DataFrameCurator:
                 using_key=self._using_key,
                 validated_only=validated_only,
                 source=self._sources.get(categorical),
+                exclude=self._exclude.get(categorical),
                 **kwargs,
             )
 
@@ -330,9 +357,11 @@ class DataFrameCurator:
 class AnnDataCurator(DataFrameCurator):
     """Curation flow for ``AnnData``.
 
-    See also :class:`~lamindb.Curate`.
+    See also :class:`~lamindb.Curator`.
+
+    Note that if genes are removed from the AnnData object, the object should be recreated using :meth:`~lamindb.Curator.from_anndata`.
 
-    Note that if genes are removed from the AnnData object, the object should be recreated using :meth:`~lamindb.Curate.from_anndata`.
+    See :doc:`docs:cellxgene-curate` for instructions on how to curate against a specific cellxgene schema version.
 
     Args:
         data: The AnnData object or an AnnData-like path.
@@ -346,7 +375,7 @@ class AnnDataCurator(DataFrameCurator):
 
     Examples:
         >>> import bionty as bt
-        >>> curate = ln.Curate.from_anndata(
+        >>> curate = ln.Curator.from_anndata(
         ...     adata,
         ...     var_index=bt.Gene.ensembl_gene_id,
         ...     categoricals={
@@ -437,6 +466,7 @@ class AnnDataCurator(DataFrameCurator):
             validated_only=validated_only,
             organism=organism,
             source=self._sources.get("var_index"),
+            exclude=self._exclude.get("var_index"),
         )
 
     def _update_registry_all(self, validated_only: bool = True, **kwargs):
@@ -536,10 +566,10 @@ class AnnDataCurator(DataFrameCurator):
 class MuDataCurator:
     """Curation flow for a ``MuData`` object.
 
-    See also :class:`~lamindb.Curate`.
+    See also :class:`~lamindb.Curator`.
 
     Note that if genes or other measurements are removed from the MuData object,
-    the object should be recreated using :meth:`~lamindb.Curate.from_mudata`.
+    the object should be recreated using :meth:`~lamindb.Curator.from_mudata`.
 
     Args:
         mdata: The MuData object to curate.
@@ -556,7 +586,7 @@ class MuDataCurator:
 
     Examples:
         >>> import bionty as bt
-        >>> curate = ln.Curate.from_mudata(
+        >>> curate = ln.Curator.from_mudata(
         ...     mdata,
         ...     var_index={
         ...         "rna": bt.Gene.ensembl_gene_id,
@@ -641,6 +671,7 @@ class MuDataCurator:
             validated_only=validated_only,
             dtype="number",
             source=self._sources.get(modality, {}).get("var_index"),
+            exclude=self._exclude.get(modality, {}).get("var_index"),
             **kwargs,
         )
 
@@ -704,6 +735,7 @@ class MuDataCurator:
             validated_only=False,
             df=self._mdata[modality].obs,
             source=self._sources.get(modality, {}).get("columns"),
+            exclude=self._exclude.get(modality, {}).get("columns"),
             **self._kwargs,  # type: ignore
             **kwargs,
         )
@@ -789,7 +821,8 @@ class MuDataCurator:
                 field=var_field,
                 key=f"{modality}_var_index",
                 using_key=self._using_key,
-                exclude=self._exclude.get(f"{modality}_var_index"),
+                source=self._sources.get(modality, {}).get("var_index"),
+                exclude=self._exclude.get(modality, {}).get("var_index"),
                 **self._kwargs,  # type: ignore
             )
             validated_var &= is_validated_var
@@ -846,19 +879,19 @@ class MuDataCurator:
         return self._artifact
 
 
-class Curate:
-    """Curation flow.
+class Curator(BaseCurator):
+    """Dataset curator.
 
     Data curation entails accurately labeling datasets with standardized metadata
     to facilitate data integration, interpretation and analysis.
 
     The curation flow has several steps:
 
-    1. Create a :class:`Curate` object corresponding to the object type that you want to curate:
+    1. Instantiate `Curator` from one of the following dataset objects:
 
-    - :meth:`~lamindb.Curate.from_df`
-    - :meth:`~lamindb.Curate.from_anndata`
-    - :meth:`~lamindb.Curate.from_mudata`
+    - :meth:`~lamindb.Curator.from_df`
+    - :meth:`~lamindb.Curator.from_anndata`
+    - :meth:`~lamindb.Curator.from_mudata`
 
     During object creation, any passed categoricals found in the object will be saved.
 
@@ -867,7 +900,7 @@ class Curate:
     - Values that can successfully validated and already exist in the registry.
     - Values which are new and not yet validated or potentially problematic values.
 
-    3. Determine how to handle validated and unvalidated values:
+    3. Determine how to handle validated and non-validated values:
 
     - Validated values not yet in the registry can be automatically registered using :meth:`~lamindb.core.DataFrameCurator.add_validated_from`.
     - Valid and new values can be registered using :meth:`~lamindb.core.DataFrameCurator.add_new_from`.
@@ -982,10 +1015,22 @@ def standardize_and_inspect(
     field: FieldAttr,
     registry: type[Record],
     standardize: bool = False,
+    exclude: str | list | None = None,
     **kwargs,
 ):
     """Standardize and inspect values using a registry."""
-    filter_kwargs = get_current_filter_kwargs(registry, kwargs)
+    # inspect exclude values in the default instance
+    values = list(values)
+    include_validated = []
+    if exclude is not None:
+        exclude = [exclude] if isinstance(exclude, str) else exclude
+        exclude = [i for i in exclude if i in values]
+        if len(exclude) > 0:
+            # exclude values are validated without source and organism
+            inspect_result_exclude = registry.inspect(exclude, field=field, mute=True)
+            # if exclude values are validated, remove them from the values
+            values = [i for i in values if i not in inspect_result_exclude.validated]
+            include_validated = inspect_result_exclude.validated
 
     if standardize:
         if hasattr(registry, "standardize") and hasattr(
@@ -993,11 +1038,17 @@ def standardize_and_inspect(
             "synonyms",  # https://github.com/laminlabs/lamindb/issues/1685
         ):
             standardized_values = registry.standardize(
-                values, field=field, mute=True, **filter_kwargs
+                values, field=field, mute=True, **kwargs
             )
             values = standardized_values
 
-    return registry.inspect(values, field=field, mute=True, **filter_kwargs)
+    inspect_result = registry.inspect(values, field=field, mute=True, **kwargs)
+    inspect_result._validated += include_validated
+    inspect_result._non_validated = [
+        i for i in inspect_result.non_validated if i not in include_validated
+    ]
+
+    return inspect_result
 
 
 def check_registry_organism(registry: Record, organism: str | None = None) -> dict:
@@ -1049,35 +1100,32 @@ def validate_categories(
         logger.indent = "   "
 
     registry = field.field.model
+
     kwargs = check_registry_organism(registry, organism)
     kwargs.update({"source": source} if source else {})
+    kwargs_current = get_current_filter_kwargs(registry, kwargs)
 
     # inspect the default instance
-    if exclude is not None:
-        exclude = [exclude] if isinstance(exclude, str) else exclude
-        # exclude values are validated without source and organism
-        inspect_result = registry.inspect(exclude, field=field, mute=True)
-        # if exclude values are validated, remove them from the values
-        values = [i for i in values if i not in inspect_result.validated]
-
     inspect_result = standardize_and_inspect(
         values=values,
         field=field,
         registry=registry,
         standardize=standardize,
-        **kwargs,
+        exclude=exclude,
+        **kwargs_current,
     )
     non_validated = inspect_result.non_validated
 
+    # inspect the using instance
     values_validated = []
     if using_key is not None and using_key != "default" and non_validated:
         registry_using = get_registry_instance(registry, using_key)
-        # inspect the using instance
         inspect_result = standardize_and_inspect(
             values=non_validated,
             field=field,
             registry=registry_using,
             standardize=standardize,
+            exclude=exclude,
             **kwargs,
         )
         non_validated = inspect_result.non_validated
@@ -1091,7 +1139,7 @@ def validate_categories(
             public_records = registry.from_values(
                 non_validated,
                 field=field,
-                **get_current_filter_kwargs(registry, kwargs),
+                **kwargs_current,
             )
             values_validated += [getattr(r, field.field.name) for r in public_records]
         finally:
@@ -1111,9 +1159,13 @@ def validate_categories(
     non_validated = [i for i in non_validated if i not in values_validated]
     n_non_validated = len(non_validated)
     if n_non_validated == 0:
-        logger.indent = ""
-        logger.success(f"{key} is validated against {colors.italic(model_field)}")
-        return True, []
+        if n_validated == 0:
+            logger.indent = ""
+            logger.success(f"{key} is validated against {colors.italic(model_field)}")
+            return True, []
+        else:
+            # validated values still need to be saved to the current instance
+            return False, []
     else:
         are = "are" if n_non_validated > 1 else "is"
         print_values = _print_values(non_validated)
@@ -1138,6 +1190,9 @@ def validate_categories_in_df(
     **kwargs,
 ) -> tuple[bool, dict]:
     """Validate categories in DataFrame columns using LaminDB registries."""
+    if not fields:
+        return True, {}
+
     if sources is None:
         sources = {}
     validated = True
@@ -1270,6 +1325,7 @@ def update_registry(
     source: Record | None = None,
     standardize: bool = True,
     warning: bool = True,
+    exclude: str | list | None = None,
     **kwargs,
 ) -> None:
     """Save features or labels records in the default instance from the using_key instance.
@@ -1329,7 +1385,8 @@ def update_registry(
             field=field,
             registry=registry,
             standardize=standardize,
-            **filter_kwargs,
+            exclude=exclude,
+            **filter_kwargs_current,
         )
         if not inspect_result_current.non_validated:
             all_labels = registry.from_values(
@@ -1348,6 +1405,7 @@ def update_registry(
             inspect_result_current.non_validated,
             field=field,
             using_key=using_key,
+            exclude=exclude,
             **filter_kwargs,
         )
 
@@ -1467,6 +1525,7 @@ def update_registry_from_using_instance(
     field: FieldAttr,
     using_key: str | None = None,
     standardize: bool = False,
+    exclude: str | list | None = None,
     **kwargs,
 ) -> tuple[list[str], list[str]]:
     """Save features or labels records from the using_key instance.
@@ -1492,6 +1551,7 @@ def update_registry_from_using_instance(
             field=field,
             registry=registry_using,
             standardize=standardize,
+            exclude=exclude,
             **kwargs,
         )
         labels_using = registry_using.filter(
@@ -1519,3 +1579,6 @@ def _save_organism(name: str):  # pragma: no cover
             )
         organism.save()
     return organism
+
+
+Curate = Curator  # backward compat

lamindb/_filter.py

@@ -1,11 +1,13 @@
 from __future__ import annotations
 
-from lnschema_core import Artifact, Collection, Record
-from lnschema_core.types import VisibilityChoice
+from typing import TYPE_CHECKING
 
-from lamindb import settings
+from lnschema_core import Artifact, Collection
 
-from ._query_set import QuerySet
+from ._query_set import QuerySet, process_expressions
+
+if TYPE_CHECKING:
+    from lnschema_core import Record
 
 
 def filter(registry: type[Record], **expressions) -> QuerySet:
@@ -13,23 +15,7 @@ def filter(registry: type[Record], **expressions) -> QuerySet:
     _using_key = None
     if "_using_key" in expressions:
         _using_key = expressions.pop("_using_key")
-    if registry in {Artifact, Collection}:
-        # visibility is set to 0 unless expressions contains id or uid equality
-        if not (
-            "id" in expressions
-            or "uid" in expressions
-            or "uid__startswith" in expressions
-        ):
-            visibility = "visibility"
-            if not any(e.startswith(visibility) for e in expressions):
-                expressions[visibility] = (
-                    VisibilityChoice.default.value
-                )  # default visibility
-            # if visibility is None, do not apply a filter
-            # otherwise, it would mean filtering for NULL values, which doesn't make
-            # sense for a non-NULLABLE column
-            elif visibility in expressions and expressions[visibility] is None:
-                expressions.pop(visibility)
+    expressions = process_expressions(registry, expressions)
     qs = QuerySet(model=registry, using=_using_key)
     if len(expressions) > 0:
         return qs.filter(**expressions)

lamindb/_finish.py

@@ -52,7 +52,7 @@ def save_context_core(
             return None
         notebook_content = read_notebook(filepath)  # type: ignore
         is_consecutive = check_consecutiveness(
-            notebook_content, calling_statement="ln.finish()"
+            notebook_content, calling_statement=".finish()"
         )
         if not is_consecutive:
             msg = "   Do you still want to proceed with finishing? (y/n) "

lamindb/_query_set.py

@@ -6,6 +6,7 @@ from typing import TYPE_CHECKING, Iterable, NamedTuple
 import pandas as pd
 from django.db import models
 from django.db.models import F
+from lamin_utils import logger
 from lamindb_setup.core._docs import doc_args
 from lnschema_core.models import (
     Artifact,
@@ -13,8 +14,10 @@ from lnschema_core.models import (
     Collection,
     IsVersioned,
     Record,
+    Registry,
     Run,
     Transform,
+    VisibilityChoice,
 )
 
 from lamindb.core.exceptions import DoesNotExist
@@ -64,6 +67,27 @@ def one_helper(self):
         return self[0]
 
 
+def process_expressions(registry: Registry, expressions: dict) -> dict:
+    if registry in {Artifact, Collection}:
+        # visibility is set to 0 unless expressions contains id or uid equality
+        if not (
+            "id" in expressions
+            or "uid" in expressions
+            or "uid__startswith" in expressions
+        ):
+            visibility = "visibility"
+            if not any(e.startswith(visibility) for e in expressions):
+                expressions[visibility] = (
+                    VisibilityChoice.default.value
+                )  # default visibility
+            # if visibility is None, do not apply a filter
+            # otherwise, it would mean filtering for NULL values, which doesn't make
+            # sense for a non-NULLABLE column
+            elif visibility in expressions and expressions[visibility] is None:
+                expressions.pop(visibility)
+    return expressions
+
+
 def get(
     registry_or_queryset: type[Record] | QuerySet,
     idlike: int | str | None = None,
@@ -88,7 +112,7 @@ def get(
             return qs.one()
     else:
         assert idlike is None  # noqa: S101
-        # below behaves exactly like `.one()`
+        expressions = process_expressions(registry, expressions)
         return registry.objects.get(**expressions)
 
 
@@ -221,6 +245,7 @@ class QuerySet(models.QuerySet, CanValidate):
         # both Transform & Run might reference artifacts
         if self.model in {Artifact, Collection, Transform, Run}:
             for record in self:
+                logger.important(f"deleting {record}")
                 record.delete(*args, **kwargs)
         else:
             self._delete_base_class(*args, **kwargs)

lamindb/core/__init__.py

@@ -30,6 +30,7 @@ Curators:
 .. autosummary::
    :toctree: .
 
+   BaseCurator
    DataFrameCurator
    AnnDataCurator
    MuDataCurator
@@ -80,6 +81,7 @@ from lnschema_core.models import (
 
 from lamindb._curate import (
     AnnDataCurator,
+    BaseCurator,
     CurateLookup,
     DataFrameCurator,
     MuDataCurator,

lamindb/core/_context.py

@@ -18,7 +18,7 @@ from ._sync_git import get_transform_reference_from_git_repo
 from ._track_environment import track_environment
 from .exceptions import (
     MissingContext,
-    NotebookNotSaved,
+    NotebookFileNotSavedToDisk,
     NotebookNotSavedError,
     NoTitleError,
     TrackNotCalled,
@@ -504,8 +504,8 @@ class Context:
                 get_seconds_since_modified(context._path) > 3
                 and os.getenv("LAMIN_TESTING") is None
             ):
-                raise NotebookNotSaved(
-                    "Please save the notebook in your editor right before running `ln.finish()`"
+                raise NotebookFileNotSavedToDisk(
+                    "Please save the notebook manually in your editor right before running `ln.finish()`"
                 )
         save_context_core(
             run=context.run,

lamindb/core/_mapped_collection.py

@@ -17,6 +17,7 @@ from .storage._anndata_accessor import (
     GroupTypes,
     StorageType,
     _safer_read_index,
+    get_spec,
     registry,
 )
 
@@ -153,13 +154,30 @@ class MappedCollection:
         self._make_connections(path_list, parallel)
 
         self.n_obs_list = []
-        for storage in self.storages:
+        for i, storage in enumerate(self.storages):
             with _Connect(storage) as store:
                 X = store["X"]
+                store_path = self.path_list[i]
+                self._check_csc_raise_error(X, "X", store_path)
                 if isinstance(X, ArrayTypes):  # type: ignore
                     self.n_obs_list.append(X.shape[0])
                 else:
                     self.n_obs_list.append(X.attrs["shape"][0])
+                for layer_key in self.layers_keys:
+                    if layer_key == "X":
+                        continue
+                    self._check_csc_raise_error(
+                        store["layers"][layer_key],
+                        f"layers/{layer_key}",
+                        store_path,
+                    )
+                if self.obsm_keys is not None:
+                    for obsm_key in self.obsm_keys:
+                        self._check_csc_raise_error(
+                            store["obsm"][obsm_key],
+                            f"obsm/{obsm_key}",
+                            store_path,
+                        )
         self.n_obs = sum(self.n_obs_list)
 
         self.indices = np.hstack([np.arange(n_obs) for n_obs in self.n_obs_list])
@@ -281,6 +299,18 @@ class MappedCollection:
         vars = pd.Index(vars)
         return [i for i, vrs in enumerate(self.var_list) if not vrs.equals(vars)]
 
+    def _check_csc_raise_error(
+        self, elem: GroupType | ArrayType, key: str, path: UPathStr
+    ):
+        if isinstance(elem, ArrayTypes):  # type: ignore
+            return
+        if get_spec(elem).encoding_type == "csc_matrix":
+            if not self.parallel:
+                self.close()
+            raise ValueError(
+                f"{key} in {path} is a csc matrix, `MappedCollection` doesn't support this format yet."
+            )
+
     def __len__(self):
         return self.n_obs
 

lamindb/core/exceptions.py

@@ -20,7 +20,7 @@ class TrackNotCalled(SystemExit):
     pass
 
 
-class NotebookNotSaved(SystemExit):
+class NotebookFileNotSavedToDisk(SystemExit):
     pass
 
 

lamindb/core/storage/__init__.py

@@ -19,7 +19,7 @@ Array accessors.
 from lamindb_setup.core.upath import LocalPathClasses, UPath, infer_filesystem
 
 from ._backed_access import AnnDataAccessor, BackedAccessor
-from ._tiledbsoma import register_for_tiledbsoma_store, write_tiledbsoma_store
+from ._tiledbsoma import save_tiledbsoma_experiment
 from ._valid_suffixes import VALID_SUFFIXES
 from .objects import infer_suffix, write_to_disk
 from .paths import delete_storage, load_to_memory

lamindb/core/storage/_tiledbsoma.py

@@ -4,16 +4,17 @@ from typing import TYPE_CHECKING, Literal
 
 from anndata import AnnData
 from lamin_utils import logger
+from lamindb_setup import settings as setup_settings
 from lamindb_setup.core._settings_storage import get_storage_region
 from lamindb_setup.core.upath import create_path
-from lnschema_core import Artifact, Run, Storage
-from upath import UPath
+from lnschema_core import Artifact, Run
 
 if TYPE_CHECKING:
     from lamindb_setup.core.types import UPathStr
     from tiledbsoma import Collection as SOMACollection
     from tiledbsoma import Experiment as SOMAExperiment
     from tiledbsoma.io import ExperimentAmbientLabelMapping
+    from upath import UPath
 
 
 def _read_adata_h5ad_zarr(objpath: UPath):
@@ -67,21 +68,44 @@ def _open_tiledbsoma(
     return SOMAType.open(storepath_str, mode=mode, context=ctx)
 
 
-def register_for_tiledbsoma_store(
-    store: UPathStr | Artifact | None,
+def save_tiledbsoma_experiment(
+    # Artifact args
     adatas: list[AnnData | UPathStr],
-    measurement_name: str,
-    obs_field_name: str,
-    var_field_name: str,
-    append_obsm_varm: bool = False,
+    key: str | None = None,
+    description: str | None = None,
     run: Run | None = None,
-) -> tuple[ExperimentAmbientLabelMapping, list[AnnData]]:
-    """Register `AnnData` objects to append to `tiledbsoma.Experiment`.
-
-    Pass the returned registration mapping and `AnnData` objects to `write_tiledbsoma_store`.
+    revises: Artifact | None = None,
+    # tiledbsoma.io.from_anndata args
+    measurement_name: str = "RNA",
+    obs_id_name: str = "obs_id",
+    var_id_name: str = "var_id",
+    append_obsm_varm: bool = False,
+    # additional keyword args for tiledbsoma.io.from_anndata
+    **kwargs,
+) -> Artifact:
+    """Write `AnnData` to `tiledbsoma.Experiment`.
 
-    See `tiledbsoma.io.from_h5ad
-    <https://tiledbsoma.readthedocs.io/en/latest/_autosummary/tiledbsoma.io.from_h5ad.html>`__.
+    Reads `AnnData` objects, writes them to `tiledbsoma.Experiment`, creates `lamindb.Artifact`
+    and saves the artifact.
+    Note that this function adds `lamin_run_uid` column to `obs` of in-memory `AnnData` objects
+    when it writes to a new store or appends to a store that has this column in `obs`.
+
+    See also `tiledbsoma.io.from_anndata
+    <https://tiledbsoma.readthedocs.io/en/latest/_autosummary/tiledbsoma.io.from_anndata.html>`__.
+
+    Args:
+        adatas: `AnnData` objects to write, in-memory or on-disk.
+        key: A relative path within default storage.
+        description: A description.
+        run: The run that creates the artifact.
+        revises: `lamindb.Artifact` with `tiledbsoma.Experiment` to append to.
+            Triggers a revision (a new untagged version).
+        measurement_name: The name of the measurement to store data in `tiledbsoma.Experiment`.
+        obs_id_name: Which `AnnData` `obs` column to use for append mode.
+        var_id_name: Which `AnnData` `var` column to use for append mode.
+        append_obsm_varm: Whether to append `obsm` and `varm` in append mode .
+        **kwargs: Additional keyword arguments passed to `tiledbsoma.io.from_anndata` that
+            writes `adatas`.
     """
     try:
         import tiledbsoma as soma
@@ -89,27 +113,38 @@ def register_for_tiledbsoma_store(
     except ImportError as e:
         raise ImportError("Please install tiledbsoma: pip install tiledbsoma") from e
 
-    if isinstance(store, Artifact):
-        storepath = store.path
+    from lamindb.core._data import get_run
+    from lamindb.core.storage.paths import auto_storage_key_from_artifact_uid
+    from lamindb.core.versioning import create_uid
+
+    run = get_run(run)
+
+    appending = revises is not None
+
+    if appending:
+        _uid = None
+        storepath = revises.path
     else:
-        storepath = None if store is None else create_path(store)
+        _uid, _ = create_uid(n_full_id=20)
+        storage_key = auto_storage_key_from_artifact_uid(
+            _uid, ".tiledbsoma", is_dir=True
+        )
+        storepath = setup_settings.storage.root / storage_key
+
+    if storepath.protocol == "s3":
+        ctx = soma.SOMATileDBContext(tiledb_config=_tiledb_config_s3(storepath))
+    else:
+        ctx = None
+
+    storepath = storepath.as_posix()
 
     add_run_uid = True
-    ctx = None
-    if storepath is not None:
-        if storepath.protocol == "s3":
-            ctx = soma.SOMATileDBContext(tiledb_config=_tiledb_config_s3(storepath))
-        if storepath.exists():
-            with soma.Experiment.open(
-                storepath.as_posix(), mode="r", context=ctx
-            ) as store:
-                add_run_uid = "lamin_run_uid" in store["obs"].schema.names
-        storepath = storepath.as_posix()
-
-    if add_run_uid:
-        from lamindb.core._data import get_run
-
-        run = get_run(run)
+    if appending:
+        with soma.Experiment.open(storepath, mode="r", context=ctx) as store:
+            add_run_uid = "lamin_run_uid" in store["obs"].schema.names
+
+    if add_run_uid and run is None:
+        raise ValueError("Pass `run`")
 
     adata_objects = []
     for adata in adatas:
@@ -117,10 +152,9 @@ def register_for_tiledbsoma_store(
             if add_run_uid:
                 if adata.is_view:
                     raise ValueError(
-                        "Can not register an `AnnData` view, please do `adata.copy()` before passing."
+                        "Can not write an `AnnData` view, please do `adata.copy()` before passing."
                     )
                 else:
-                    logger.warning("Mutating in-memory AnnData.")
                     adata.obs["lamin_run_uid"] = run.uid
         else:
             adata = _read_adata_h5ad_zarr(create_path(adata))
@@ -128,102 +162,31 @@ def register_for_tiledbsoma_store(
                 adata.obs["lamin_run_uid"] = run.uid
         adata_objects.append(adata)
 
-    registration_mapping = soma_io.register_anndatas(
-        experiment_uri=storepath,
-        adatas=adata_objects,
-        measurement_name=measurement_name,
-        obs_field_name=obs_field_name,
-        var_field_name=var_field_name,
-        append_obsm_varm=append_obsm_varm,
-        context=ctx,
-    )
-
-    return registration_mapping, adata_objects
-
-
-def write_tiledbsoma_store(
-    store: Artifact | UPathStr,
-    adata: AnnData | UPathStr,
-    run: Run | None = None,
-    artifact_kwargs: dict | None = None,
-    **kwargs,
-) -> Artifact:
-    """Write `AnnData` to `tiledbsoma.Experiment`.
-
-    Reads `AnnData`, writes it to `tiledbsoma.Experiment` and creates `lamindb.Artifact`.
-
-    See `tiledbsoma.io.from_h5ad
-    <https://tiledbsoma.readthedocs.io/en/latest/_autosummary/tiledbsoma.io.from_h5ad.html>`__.
-    """
-    try:
-        import tiledbsoma as soma
-        import tiledbsoma.io as soma_io
-    except ImportError as e:
-        raise ImportError("Please install tiledbsoma: pip install tiledbsoma") from e
-
-    from lamindb.core._data import get_run
-
-    if artifact_kwargs is None:
-        artifact_kwargs = {}
-
-    appending: bool = kwargs.get("registration_mapping", None) is not None
-    store_is_artifact: bool = isinstance(store, Artifact)
-    if store_is_artifact:
-        if not appending:
-            raise ValueError(
-                "Trying to append to an existing store without `registration_mapping`."
-            )
-        storepath = store.path
-    else:
-        storepath = create_path(store)
-    add_run_uid: bool = not appending
-
-    if not isinstance(adata, AnnData):
-        # create_path is used
-        # in case adata is somewhere in our managed s3 bucket or just in s3
-        adata = _read_adata_h5ad_zarr(create_path(adata))
-    elif add_run_uid and adata.is_view:
-        raise ValueError(
-            "Can not write from an `AnnData` view, please do `adata.copy()` before passing."
+    if appending or len(adata_objects) > 1:
+        registration_mapping = soma_io.register_anndatas(
+            experiment_uri=storepath if appending else None,
+            adatas=adata_objects,
+            measurement_name=measurement_name,
+            obs_field_name=obs_id_name,
+            var_field_name=var_id_name,
+            append_obsm_varm=append_obsm_varm,
+            context=ctx,
         )
-
-    run = get_run(run)
-
-    if add_run_uid:
-        adata.obs["lamin_run_uid"] = run.uid
-
-    if storepath.protocol == "s3":
-        ctx = soma.SOMATileDBContext(tiledb_config=_tiledb_config_s3(storepath))
     else:
-        ctx = None
-
-    soma_io.from_anndata(storepath.as_posix(), adata, context=ctx, **kwargs)
-
-    if add_run_uid:
-        del adata.obs["lamin_run_uid"]
+        registration_mapping = None
+
+    for adata_obj in adata_objects:
+        soma_io.from_anndata(
+            storepath,
+            adata_obj,
+            measurement_name,
+            context=ctx,
+            obs_id_name=obs_id_name,
+            var_id_name=var_id_name,
+            registration_mapping=registration_mapping,
+            **kwargs,
+        )
 
-    revises = None
-    if appending:
-        if store_is_artifact:
-            revises = store
-        else:
-            from lamindb._artifact import (
-                check_path_in_existing_storage,
-                get_relative_path_to_directory,
-            )
-
-            storage = check_path_in_existing_storage(storepath)
-            if isinstance(storage, Storage):
-                search_by_key = get_relative_path_to_directory(
-                    path=storepath, directory=UPath(storage.root)
-                ).as_posix()
-                revises = Artifact.filter(
-                    key=search_by_key, is_latest=True, _key_is_virtual=False
-                ).one_or_none()
-                if revises is not None:
-                    logger.info(f"Assuming it is a new version of {revises}.")
-
-    if revises is None:
-        return Artifact(storepath, run=run, **artifact_kwargs)
-    else:
-        return Artifact(storepath, run=run, revises=revises, **artifact_kwargs)
+    return Artifact(
+        storepath, key=key, description=description, run=run, revises=revises, _uid=_uid
+    ).save()

lamindb/integrations/__init__.py

@@ -4,6 +4,9 @@
    :toctree: .
 
    save_vitessce_config
+   save_tiledbsoma_experiment
 """
 
+from lamindb.core.storage import save_tiledbsoma_experiment
+
 from ._vitessce import save_vitessce_config

{lamindb-0.76.2.dist-info → lamindb-0.76.3.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: lamindb
-Version: 0.76.2
+Version: 0.76.3
 Summary: A data framework for biology.
 Author-email: Lamin Labs <open-source@lamin.ai>
 Requires-Python: >=3.8
@@ -9,8 +9,8 @@ Classifier: Programming Language :: Python :: 3.8
 Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
-Requires-Dist: lnschema_core==0.73.2
-Requires-Dist: lamindb_setup==0.76.6
+Requires-Dist: lnschema_core==0.73.3
+Requires-Dist: lamindb_setup==0.76.7
 Requires-Dist: lamin_utils==0.13.4
 Requires-Dist: lamin_cli==0.16.2
 Requires-Dist: rapidfuzz
@@ -24,7 +24,7 @@ Requires-Dist: pandas
 Requires-Dist: graphviz
 Requires-Dist: psycopg2-binary
 Requires-Dist: lamindb_setup[aws] ; extra == "aws"
-Requires-Dist: bionty==0.48.3 ; extra == "bionty"
+Requires-Dist: bionty==0.49.0 ; extra == "bionty"
 Requires-Dist: pre-commit ; extra == "dev"
 Requires-Dist: nox ; extra == "dev"
 Requires-Dist: laminci>=0.3 ; extra == "dev"

{lamindb-0.76.2.dist-info → lamindb-0.76.3.dist-info}/RECORD

@@ -1,17 +1,17 @@
-lamindb/__init__.py,sha256=OCLxYHwhRyaf6Zx-AKHTSEFWM-J342YRxDP2hhCIu6c,2309
-lamindb/_artifact.py,sha256=RygNHePDo895KG61FctFRdHO4jVa9cmj-yBBsyRE50A,43840
+lamindb/__init__.py,sha256=iZlJlG4hUmSRNZvau2bAEG2Z9B-qnWWNazYYw5m3ciM,2351
+lamindb/_artifact.py,sha256=6iD2qzZxWUHPiItUdEJt9BDbYVv6HF3bgpW2ZDjMf7s,44093
 lamindb/_can_validate.py,sha256=ne8-9sAG9vXnMXZqso6mYMt-xDg16h-gq6yHJXKFpuI,17690
 lamindb/_collection.py,sha256=F_VgpLBprrzUQ-tPngWvO9vFd7jX66MVwIi031JOris,14871
-lamindb/_curate.py,sha256=EAikoFg2rL8kRsWEIXH33P1VplSGAACMuFHanicguWw,55964
+lamindb/_curate.py,sha256=_VISRycaPaburnJKuFM5XpEWqNGozdUcwBGMj_knZpM,58041
 lamindb/_feature.py,sha256=nZhtrH0ssoNls-hV-dkwfK9sKypg2El59R9qfarxfUE,5340
 lamindb/_feature_set.py,sha256=DmAy96V_RyV0yiyvWOCHgustXPsCaMwn4TrWwh2qDd8,8104
-lamindb/_filter.py,sha256=Fs7_x3tA_KVz0lYt0bJkWfjnEKAnkVN20CfKhp9aKWg,1403
-lamindb/_finish.py,sha256=wCjPmTBmL_z2WcZq9v6TZroZ_J_Te9KC5GzFuYdrIRc,9413
+lamindb/_filter.py,sha256=9QHa9J-_6QeYPQATZpTun2VGiFofwzB0Km-KnKajHcM,663
+lamindb/_finish.py,sha256=Yytv26ruL7EFxKewOrgqJxDircZVCpHOkwUjqdh4oMY,9411
 lamindb/_from_values.py,sha256=8kYpR8Q85EOaTcsPGjVHeZh29fGVgum5OEQf4Hsz_80,13533
 lamindb/_is_versioned.py,sha256=5lAnhTboltFkZCKVRV1uxkm0OCjJz_HKi3yQq_vEuMs,1306
 lamindb/_parents.py,sha256=eMavdd6IO6STOVJSlR2TzdRtx6sKYDKsMOtlR3DZlgQ,15599
 lamindb/_query_manager.py,sha256=Ipe85HL31DDwMbC8CN_1Svbwk48a_DUh_INGQdZL08I,4222
-lamindb/_query_set.py,sha256=mwNfGWcZbI5a1VfuiQzkweU4VQchE8Va8HiCIa0GNu4,11604
+lamindb/_query_set.py,sha256=rykST1bxbk2_dt8Rx1TGTEHzXUc1yyc0rZWHO6bRGBA,12667
 lamindb/_record.py,sha256=MDLvcsZ23lZkThnrYQtj_uW-1BnxCPpyJ5mKl9KPs4A,21199
 lamindb/_run.py,sha256=5M_r1zGDv9HlqbqRKTWCYCOtENovJ-8mQ4kY7XqcLaU,1888
 lamindb/_save.py,sha256=Fu7Z84btKOXfTfpunKLni21s5ER2zIllqg5e3nPq-0A,10910
@@ -20,16 +20,16 @@ lamindb/_transform.py,sha256=Ck674iuKIp9HDpHuyFFM2xKJcFAUvQDRTnEdnx00Rq0,4114
 lamindb/_ulabel.py,sha256=XDSdZBXX_ki5s1vOths3MjF2x5DPggBR_PV_KF4SGyg,1611
 lamindb/_utils.py,sha256=LGdiW4k3GClLz65vKAVRkL6Tw-Gkx9DWAdez1jyA5bE,428
 lamindb/_view.py,sha256=4Ln2ItTb3857PAI-70O8eJYqoTJ_NNFc7E_wds6OGns,2412
-lamindb/core/__init__.py,sha256=QePKN3dkGuNEfIc2CKwBt3Kcl39bbghX6XwXKpK3BA0,1491
-lamindb/core/_context.py,sha256=O4ovdIfvIC9E4oaqpSjBqwHOKThmP8_aod0MiOWaEwI,19517
+lamindb/core/__init__.py,sha256=Nq-YbQT8X0KXoGXgctr9lm8H7J6sZeBBJ3hixGq0iuQ,1523
+lamindb/core/_context.py,sha256=FamAbmq3VZYrgwZkaRtkBY3f5YepLtlblh8wyv4a1OQ,19546
 lamindb/core/_data.py,sha256=eocOXsZGu62LPtz6yIlvHhPSJTf3yF2ITZTffyflWYI,16269
 lamindb/core/_feature_manager.py,sha256=94tX6gq_Rx7fkDARQBxB2z92qUDpHocFSAdKv5izMT4,32490
 lamindb/core/_label_manager.py,sha256=jSLvxAuTuOYosSh_QJhIz3bqnbmWKP43y5abVMb-hOQ,9240
-lamindb/core/_mapped_collection.py,sha256=ST-cTfokIGkRadjSHEyvIK2san8cGr7WZpgbgs5neLI,22025
+lamindb/core/_mapped_collection.py,sha256=1XzratL2IvRleqioNhWo26Lsuqkev8-HEImmHQxw9Kw,23266
 lamindb/core/_settings.py,sha256=GGEB8BU5GinIfD4ktr1Smp6GPHGaInu46MhP4EecZDY,5950
 lamindb/core/_sync_git.py,sha256=qc0yfPyKeG4uuNT_3qsv-mkIMqhLFqfXNeNVO49vV00,4547
 lamindb/core/_track_environment.py,sha256=STzEVUzOeUEWdX7WDJUkKH4u08k7eupRX6AXQwoVt14,828
-lamindb/core/exceptions.py,sha256=qNFYN5Jc7Y6kw4re-jsW0AEIprsV2HB1wTcJiO-u-ks,1278
+lamindb/core/exceptions.py,sha256=TKyt1JOUwWIHbkCQjir_LQadf8960eQ95RWhSpz5_Bk,1288
 lamindb/core/fields.py,sha256=47Jmh3efUr5ZscgimR_yckY-I3cNf8ScLutbwKCK3j4,162
 lamindb/core/schema.py,sha256=KiYQn_8fokSMztTNDe6qUocZzKXWxU32H-YChNJv51A,1877
 lamindb/core/types.py,sha256=uVBqSVLoQaTkqP9nqsJhwU6yYnx8H5e6-ZxrB6vpOOw,265
@@ -37,11 +37,11 @@ lamindb/core/versioning.py,sha256=__SOHhk5OjMJgAUjixzp0GFcQrpjm8sBUXC9Fouk2AE,51
 lamindb/core/datasets/__init__.py,sha256=zRP98oqUAaXhqWyKMiH0s_ImVIuNeziQQ2kQ_t0f-DI,1353
 lamindb/core/datasets/_core.py,sha256=CgVF_pXuBXLElzubDMsl1DbpYOnXCY0HleITVvBKih4,19873
 lamindb/core/datasets/_fake.py,sha256=BZF9R_1iF0HDnvtZNqL2FtsjSMuqDIfuFxnw_LJYIh4,953
-lamindb/core/storage/__init__.py,sha256=9B3JqHydQnclP4NUY2kEc99K1cJBQZA4jyy3EmDxsYk,541
+lamindb/core/storage/__init__.py,sha256=x-Bpxv1rx6uGog-chI0fdpt0UhkXQkwoQqye0TNk0WQ,514
 lamindb/core/storage/_anndata_accessor.py,sha256=jmEZeeZlt8-qBXRkU0tTA-t6dVEb_dH86wc1ok0jSRY,24030
 lamindb/core/storage/_anndata_sizes.py,sha256=aXO3OB--tF5MChenSsigW6Q-RuE8YJJOUTVukkLrv9A,1029
 lamindb/core/storage/_backed_access.py,sha256=YcWCeT2eligJGsBdjJS_-4el_eC9J088jxUWG9lsleM,3231
-lamindb/core/storage/_tiledbsoma.py,sha256=GNPG8pDYmZLFBSujsQ8VqlfWaFmDO58kgBXw6JESQJs,7812
+lamindb/core/storage/_tiledbsoma.py,sha256=_ObAT3pIx0OA5uq55v6aWnIeyTTKjetjV3Gkk4oDWR0,6851
 lamindb/core/storage/_valid_suffixes.py,sha256=vUSeQ4s01rdhD_vSd6wKmFBsgMJAKkBMnL_T9Y1znMg,501
 lamindb/core/storage/_zarr.py,sha256=5ceEz6YIvgvUnVVNWhK5Z4W0WfrvyvY82Yna5jSX1_E,3661
 lamindb/core/storage/objects.py,sha256=OzvBCS-Urz5mr-O95qYt6RGBDDX5HmjfRRKWPPDn1ZE,1797
@@ -49,11 +49,11 @@ lamindb/core/storage/paths.py,sha256=woOrjtBhNnzm8DjF262ipwyZaQ_A-7MT2ZPoiefAfYk
 lamindb/core/subsettings/__init__.py,sha256=KFHPzIE7f7Bj4RgMjGQF4CjTdHVG_VNFBrCndo49ixo,198
 lamindb/core/subsettings/_creation_settings.py,sha256=54mfMH_osC753hpxcl7Dq1rwBD2LHnWveXtQpkLBITE,1194
 lamindb/core/subsettings/_transform_settings.py,sha256=4YbCuZtJo6zdytl6UQR4GvdDkTtT6SRBqVzofGzNOt8,583
-lamindb/integrations/__init__.py,sha256=MoLRD_qqX5WHFUAqHL6zoY_cDkWH0zimaGT_1CyXKnk,124
+lamindb/integrations/__init__.py,sha256=RWGMYYIzr8zvmNPyVB4m-p4gMDhxdRbjES2Ed23OItw,215
 lamindb/integrations/_vitessce.py,sha256=aDCyZTddpMfUzjEo0DXQ3XlD--ebSqnsGiMxJBunX90,5141
 lamindb/setup/__init__.py,sha256=OwZpZzPDv5lPPGXZP7-zK6UdO4FHvvuBh439yZvIp3A,410
 lamindb/setup/core/__init__.py,sha256=SevlVrc2AZWL3uALbE5sopxBnIZPWZ1IB0NBDudiAL8,167
-lamindb-0.76.2.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-lamindb-0.76.2.dist-info/WHEEL,sha256=EZbGkh7Ie4PoZfRQ8I0ZuP9VklN_TvcZ6DSE5Uar4z4,81
-lamindb-0.76.2.dist-info/METADATA,sha256=99nkmjVubVfQahE9S4aO1e8Ot6G4iFaPZltNBE-oFPo,2381
-lamindb-0.76.2.dist-info/RECORD,,
+lamindb-0.76.3.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+lamindb-0.76.3.dist-info/WHEEL,sha256=EZbGkh7Ie4PoZfRQ8I0ZuP9VklN_TvcZ6DSE5Uar4z4,81
+lamindb-0.76.3.dist-info/METADATA,sha256=DoLPrLaR-VmA_7wyT_AM-fLUqi-0qL9OhxKT4U-wvWU,2381
+lamindb-0.76.3.dist-info/RECORD,,