DataTypes

laktory.models.dtypes.DType

Bases: BaseModel

Generic data type class.

Examples:

from laktory import models

# Int32
dtype = models.DType(name="Int32")
print(dtype)

# List of string
dtype = models.DType(name="list", inner="str")
print(dtype)

# Structure
dtype = models.DType(
    name="struct", fields={"x": {"name": "list", "inner": "double"}, "y": dtype}
)
print(dtype)

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.DField

Bases: BaseModel

Data Field definition

PARAMETER	DESCRIPTION
dtype	Field data type TYPE: str | DType | VariableType
name	Field name TYPE: str | VariableType

---

laktory.models.dtypes.Array

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType
name	Data type name. TYPE: str | VariableType DEFAULT: 'Array'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.List

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType
name	Data type name. TYPE: str | VariableType DEFAULT: 'List'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.Struct

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'Struct'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.Decimal

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'Decimal'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.Int128

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'Int128'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.Int64

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'Int64'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.Int32

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'Int32'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.Int16

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'Int16'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.Int8

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'Int8'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.UInt128

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'UInt128'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.UInt64

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'UInt64'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.UInt32

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'UInt32'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.UInt16

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'UInt16'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.UInt8

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'UInt8'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.Float32

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'Float32'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.Float64

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'Float64'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.String

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'String'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.Boolean

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'Boolean'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.Object

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'Object'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.Categorical

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'Categorical'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.Enum

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'Enum'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.Datetime

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'Datetime'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.Duration

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'Duration'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )

---

laktory.models.dtypes.Date

Bases: SpecificDType

PARAMETER	DESCRIPTION
category	Data type category TYPE: Literal['NUMERIC', 'STRING', 'STRUCT'] | VariableType DEFAULT: 'NUMERIC'
fields	Definition of fields for Struct type. TYPE: list[DField | VariableType] | VariableType DEFAULT: None
inner	Data type for sub-elements for Array or List types. TYPE: str | DType | VariableType DEFAULT: None
name	Data type name. TYPE: str | VariableType DEFAULT: 'Date'
shape	Definition of shape for Array type. TYPE: int | list[int] | VariableType DEFAULT: None
time_unit	Time unit for Datetime or Duration types (e.g., 'us', 'ms', 'ns'). TYPE: str | None | VariableType DEFAULT: None
time_zone	Time zone for Datetime type (e.g., 'UTC', 'America/New_York'). Set to None for timezone-naive (TimestampNTZ in Spark). TYPE: str | None | VariableType DEFAULT: 'UTC'

METHOD	DESCRIPTION
set_name	Force dump of name property to allow de-serialization from DType model
to_narwhals	Get equivalent Narwhals data type
to_polars	Get equivalent Polars data type
to_pyspark	Get equivalent Spark data type

set_name()

Force dump of name property to allow de-serialization from DType model

Source code in laktory/models/dtypes.py

@model_validator(mode="after")
def set_name(self) -> Any:
    """Force dump of `name` property to allow de-serialization from DType model"""
    self.model_fields_set.add("name")
    return self

to_narwhals()

Get equivalent Narwhals data type

Source code in laktory/models/dtypes.py

def to_narwhals(self):
    """Get equivalent Narwhals data type"""
    nw_dtypes = nw.dtypes
    _type = self.name

    # Complex types
    if _type in "Array":
        return nw_dtypes.Array(inner=self.inner.to_narwhals(), shape=self.shape)

    if _type == "List":
        return nw_dtypes.List(inner=self.inner.to_narwhals())

    if _type == "Struct":
        fields = []
        for field in self.fields:
            fields += [nw.Field(name=field.name, dtype=field.dtype.to_narwhals())]
        return nw_dtypes.Struct(fields)

    if _type == "Datetime":
        return nw_dtypes.Datetime(
            time_unit=self.time_unit or "us",
            time_zone=self.time_zone,
        )

    if _type == "Duration":
        return nw_dtypes.Duration(time_unit=self.time_unit or "us")

    if hasattr(nw_dtypes, _type):
        return getattr(nw_dtypes, _type)()

    # Not Found
    raise ValueError(f"Data type with name '{self.name}' is not supported")

to_polars()

Get equivalent Polars data type

Source code in laktory/models/dtypes.py

def to_polars(self):
    """Get equivalent Polars data type"""
    from narwhals._polars.utils import narwhals_to_native_dtype

    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
    )

to_pyspark()

Get equivalent Spark data type

Source code in laktory/models/dtypes.py

def to_pyspark(self):
    """Get equivalent Spark data type"""
    import pyspark.sql.types as T
    from narwhals._spark_like.utils import narwhals_to_native_dtype

    from laktory import get_spark_session

    # Narwhals rejects time zones that differ from the session's, so bypass
    # it for Datetime: None → TimestampNTZType (explicit TZ-naive),
    # any time zone → TimestampType (timezone-aware).
    if self.name == "Datetime":
        return T.TimestampNTZType() if self.time_zone is None else T.TimestampType()

    spark = get_spark_session()
    return narwhals_to_native_dtype(
        dtype=self.to_narwhals(),
        version=nw._utils.Version.MAIN,
        spark_types=T,
        session=spark,
    )