mendevi.database.serialize¶

Serialize the sql binary data.

Functions

`binary_to_list`(data)	Deserialize a binary data into a float vector.
`binary_to_tensor`(data)	Serialize a 2d array into binary data.
`list_to_binary`(vect)	Serialize a float vector into binary data.
`skip_none`(func)	Allow skipping the process of None input.
`tensor_to_binary`(tensor)	Serialize a 2d array into binary data.

Details

mendevi.database.serialize.binary_to_list(data: bytes) → ndarray[float32][source]

Deserialize a binary data into a float vector.

Bijection of list_to_binary().

Parameters¶

databytes: The serialized float 32 list.

Returns¶

vectnp.ndarray: The list of floats

Examples¶

>>> from mendevi.database.serialize import binary_to_list
>>> binary_to_list(b'\x00\x00\x00\x00\x00\x00\x80\xbf\xea\xd6\xfc=')
array([ 0.        , -1.        ,  0.12345679], dtype=float32)
>>>

mendevi.database.serialize.binary_to_tensor(data: bytes) → ndarray[float32, float32][source]

Serialize a 2d array into binary data.

Bijection os binary_to_tensor().

Parameters¶

databytes: The serialized float32 tensor.

Returns¶

tensornp.ndarray: The list of list of float

Examples¶

>>> from mendevi.database.serialize import binary_to_tensor
>>> data = b'\x02\x00\x00\x00\xcd\xcc\x8c?\xcd\xcc\x0c@33S@\xcd\xcc\x8c@\x00\x00\xb0@33\xd3@'
>>> binary_to_tensor(data)
array([[1.1, 2.2],
       [3.3, 4.4],
       [5.5, 6.6]], dtype=float32)
>>>

mendevi.database.serialize.list_to_binary(vect: list[float]) → bytes[source]

Serialize a float vector into binary data.

Bijection of binary_to_list().

Parameters¶

vectarralike: The list of floats

Returns¶

databytes: The serialized float32 list.

Examples¶

>>> from mendevi.database.serialize import list_to_binary
>>> list_to_binary([0.0, -1.0, 0.123456789])
b'\x00\x00\x00\x00\x00\x00\x80\xbf\xea\xd6\xfc='
>>>

mendevi.database.serialize.skip_none(func: Callable) → Callable[source]: Allow skipping the process of None input.

mendevi.database.serialize.tensor_to_binary(tensor: list[list[float]]) → bytes[source]

Serialize a 2d array into binary data.

Bijection os binary_to_tensor().

Parameters¶

tensorarraylike: The list of list of float

Returns¶

databytes: The serialized float32 tensor.

Examples¶

>>> from mendevi.database.serialize import tensor_to_binary
>>> tensor_to_binary([[1.1, 2.2], [3.3, 4.4], [5.5, 6.6]])
b'\x02\x00\x00\x00\xcd\xcc\x8c?\xcd\xcc\x0c@33S@\xcd\xcc\x8c@\x00\x00\xb0@33\xd3@'
>>>