pyflink.dataframe.DataType

class DataType(_table_type: pyflink.table.types.DataType)

DataType for DataFrame API.

This class provides a factory method style for creating data types. It wraps the underlying Flink Table API DataType for compatibility.

Example::

>>> from pyflink.dataframe import DataType
>>> DataType.int32()
>>> DataType.int64()
>>> DataType.string()
>>> DataType.list(DataType.int32())
>>> DataType.struct({"name": DataType.string(), "age": DataType.int32()})

Methods

binary()	Binary type (BYTES).
bool()	Boolean type.
boolean()	Boolean type.
date()	Date type (year, month, day).
decimal(precision, scale)	Decimal type with specified precision and scale.
fixed_size_binary(length)	Fixed-length binary type (BINARY(n)).
fixed_size_string(length)	Fixed-length character type (CHAR(n)).
float32()	32-bit floating point type (FLOAT).
float64()	64-bit floating point type (DOUBLE).
image([mode, height, width])	Decoded image type.
int16()	16-bit signed integer type (SMALLINT).
int32()	32-bit signed integer type (INT).
int64()	64-bit signed integer type (BIGINT).
int8()	8-bit signed integer type (TINYINT).
list(dtype)	List type with element type.
map(key_type, value_type)	Map type with key and value types.
not_null()	Return a non-nullable version of this type.
null()	Null type (NULL).
nullable()	Return a nullable version of this type.
row(fields)	Struct/Row type with named fields.
string()	Variable-length string type (VARCHAR).
struct(fields)	Struct/Row type with named fields.
tensor(dtype, shape)	Fixed-shape tensor type.
time([precision])	Time type (hour, minute, second, fractional seconds).
timestamp([precision])	Timestamp type without time zone.
timestamp_ltz([precision])	Timestamp type with local time zone.
variant()	Variant type that can store a value of any type.