lamindb 1.3.0__py3-none-any.whl → 1.3.2__py3-none-any.whl

lamindb/__init__.py

@@ -72,7 +72,7 @@ Backward compatibility.
 
 # ruff: noqa: I001
 # denote a release candidate for 0.1.0 with 0.1rc1, 0.1a1, 0.1b1, etc.
-__version__ = "1.3.0"
+__version__ = "1.3.2"
 
 import warnings
 

lamindb/_view.py

@@ -11,7 +11,7 @@ from lamindb_setup._init_instance import get_schema_module_name
 
 from lamindb.models import Feature, FeatureValue, ParamValue, Record
 
-from .models.feature import convert_pandas_dtype_to_lamin_dtype
+from .models.feature import serialize_pandas_dtype
 
 if TYPE_CHECKING:
     import pandas as pd
@@ -114,7 +114,7 @@ def view(
     """
     if df is not None:
         descriptions = {
-            col_name: convert_pandas_dtype_to_lamin_dtype(dtype)
+            col_name: serialize_pandas_dtype(dtype)
             for col_name, dtype in df.dtypes.to_dict().items()
         }
         feature_dtypes = dict(Feature.objects.values_list("name", "dtype"))

lamindb/base/types.py

@@ -7,7 +7,7 @@ Central object types.
 
    ArtifactKind
    TransformType
-   FeatureDtype
+   Dtype
 
 Basic types.
 
@@ -38,14 +38,53 @@ TransformType = Literal[
     "pipeline", "notebook", "upload", "script", "function", "linker"
 ]
 ArtifactKind = Literal["dataset", "model"]
-FeatureDtype = Literal[
-    "cat",  # categorical variables
-    "num",  # numerical variables
-    "str",  # string variables
-    "int",  # integer variables
-    "float",  # float variables
-    "bool",  # boolean variables
-    "date",  # date variables
-    "datetime",  # datetime variables
-    "object",  # this is a pandas type, we're only using it for complicated types, not for strings
+
+# below is used for Feature.dtype and Param.dtype
+Dtype = Literal[
+    "cat",  # categoricals
+    "num",  # numericals
+    "str",  # string
+    "int",  # integer / numpy.integer
+    "float",  # float
+    "bool",  # boolean
+    "date",  # date
+    "datetime",  # datetime
+    "object",  # this is a pandas input dtype, we're only using it for complicated types, not for strings
 ]
+"""Data type.
+
+Data types in lamindb are a string-serialized abstraction of common data types.
+
+Overview
+========
+
+============  ============  =================================================
+description   lamindb       pandas
+============  ============  =================================================
+categorical   `"cat"`       `category`
+numerical     `"num"`       `int | float`
+integer       `"int"`       `int64 | int32 | int16 | int8 | uint | ...`
+float         `"float"`     `float64 | float32 | float16 | float8 | ...`
+string        `"str"`       `object`
+datetime      `"datetime"`  `datetime`
+date          `"date"`      `date`
+============  ============  =================================================
+
+Categoricals
+============
+
+Beyond indicating that a feature is a categorical, `lamindb` allows you to define the registry to which values are restricted.
+
+For example, `'cat[ULabel]'` or `'cat[bionty.CellType]'` indicate that permissible values are from the `ULabel` or `CellType` registry, respectively.
+
+You can also reference multiple registries, e.g., `'cat[ULabel|bionty.CellType]'` indicates that values can be from either registry.
+
+You can also restrict to sub-types defined in registries via the `type` column, e.g., `'cat[ULabel[CellMedium]]'` indicates that values must be of type `CellMedium` within the `ULabel` registry.
+
+Literal
+=======
+
+A `Dtype` object in `lamindb` is a `Literal` up to further specification of `"cat"`.
+
+"""
+FeatureDtype = Dtype  # backward compat

lamindb/core/types.py

@@ -6,7 +6,7 @@ from anndata import AnnData
 from lamindb_setup.core.types import UPathStr
 
 from lamindb.base.types import (
-    FeatureDtype,
+    Dtype,
     FieldAttr,
     ListLike,
     StrField,