tf.keras.layers.experimental.preprocessing.Hashing
Stay organized with collections
Save and categorize content based on your preferences.
Implements categorical feature hashing, also known as "hashing trick".
Inherits From: Layer
tf.keras.layers.experimental.preprocessing.Hashing(
num_bins, salt=None, name=None, **kwargs
)
This layer transforms single or multiple categorical inputs to hashed output.
It converts a sequence of int or string to a sequence of int. The stable hash
function uses tensorflow::ops::Fingerprint to produce universal output that
is consistent across platforms.
This layer uses FarmHash64 by default,
which provides a consistent hashed output across different platforms and is
stable across invocations, regardless of device and context, by mixing the
input bits thoroughly.
If you want to obfuscate the hashed output, you can also pass a random salt
argument in the constructor. In that case, the layer will use the
SipHash64 hash function, with
the salt value serving as additional input to the hash function.
Example (FarmHash64):
layer = tf.keras.layers.experimental.preprocessing.Hashing(num_bins=3)
inp = [['A'], ['B'], ['C'], ['D'], ['E']]
layer(inp)
<tf.Tensor: shape=(5, 1), dtype=int64, numpy=
array([[1],
[0],
[1],
[1],
[2]])>
Example (FarmHash64) with list of inputs:
layer = tf.keras.layers.experimental.preprocessing.Hashing(num_bins=3)
inp_1 = [['A'], ['B'], ['C'], ['D'], ['E']]
inp_2 = np.asarray([[5], [4], [3], [2], [1]])
layer([inp_1, inp_2])
Example (SipHash64):
layer = tf.keras.layers.experimental.preprocessing.Hashing(num_bins=3,
salt=[133, 137])
inp = [['A'], ['B'], ['C'], ['D'], ['E']]
layer(inp)
<tf.Tensor: shape=(5, 1), dtype=int64, numpy=
array([[1],
[2],
[1],
[0],
[2]])>
Example (Siphash64 with a single integer, same as salt=[133, 133]
layer = tf.keras.layers.experimental.preprocessing.Hashing(num_bins=3,
salt=133)
inp = [['A'], ['B'], ['C'], ['D'], ['E']]
layer(inp)
<tf.Tensor: shape=(5, 1), dtype=int64, numpy=
array([[0],
[0],
[2],
[1],
[0]])>
Reference: SipHash with salt
Arguments |
|---|
num_bins
|
Number of hash bins.
|
salt
|
A single unsigned integer or None.
If passed, the hash function used will be SipHash64, with these values
used as an additional input (known as a "salt" in cryptography).
These should be non-zero. Defaults to None (in that
case, the FarmHash64 hash function is used). It also supports
tuple/list of 2 unsigned integer numbers, see reference paper for details.
|
name
|
Name to give to the layer.
|
**kwargs
|
Keyword arguments to construct a layer.
|
Input shape: A single or list of string, int32 or int64 Tensor,
SparseTensor or RaggedTensor of shape [batch_size, ...,]
Output shape: An int64 Tensor, SparseTensor or RaggedTensor of shape
[batch_size, ...]. If any input is RaggedTensor then output is
RaggedTensor, otherwise if any input is SparseTensor then output is
SparseTensor, otherwise the output is Tensor.
Except as otherwise noted, the content of this page is licensed under the Creative Commons Attribution 4.0 License, and code samples are licensed under the Apache 2.0 License. For details, see the Google Developers Site Policies. Java is a registered trademark of Oracle and/or its affiliates.
Last updated 2020-10-01 UTC.
[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Missing the information I need","missingTheInformationINeed","thumb-down"],["Too complicated / too many steps","tooComplicatedTooManySteps","thumb-down"],["Out of date","outOfDate","thumb-down"],["Samples / code issue","samplesCodeIssue","thumb-down"],["Other","otherDown","thumb-down"]],["Last updated 2020-10-01 UTC."],[],[],null,["# tf.keras.layers.experimental.preprocessing.Hashing\n\n\u003cbr /\u003e\n\n|------------------------------------------------------------------------------------------------------------------------------------------------|\n| [View source on GitHub](https://github.com/tensorflow/tensorflow/blob/v2.3.0/tensorflow/python/keras/layers/preprocessing/hashing.py#L43-L280) |\n\nImplements categorical feature hashing, also known as \"hashing trick\".\n\nInherits From: [`Layer`](../../../../../tf/keras/layers/Layer)\n\n#### View aliases\n\n\n**Compat aliases for migration**\n\nSee\n[Migration guide](https://www.tensorflow.org/guide/migrate) for\nmore details.\n\n[`tf.compat.v1.keras.layers.experimental.preprocessing.Hashing`](/api_docs/python/tf/keras/layers/experimental/preprocessing/Hashing)\n\n\u003cbr /\u003e\n\n tf.keras.layers.experimental.preprocessing.Hashing(\n num_bins, salt=None, name=None, **kwargs\n )\n\nThis layer transforms single or multiple categorical inputs to hashed output.\nIt converts a sequence of int or string to a sequence of int. The stable hash\nfunction uses tensorflow::ops::Fingerprint to produce universal output that\nis consistent across platforms.\n\nThis layer uses [FarmHash64](https://github.com/google/farmhash) by default,\nwhich provides a consistent hashed output across different platforms and is\nstable across invocations, regardless of device and context, by mixing the\ninput bits thoroughly.\n\nIf you want to obfuscate the hashed output, you can also pass a random `salt`\nargument in the constructor. In that case, the layer will use the\n[SipHash64](https://github.com/google/highwayhash) hash function, with\nthe `salt` value serving as additional input to the hash function.\n\nExample (FarmHash64): \n\n layer = tf.keras.layers.experimental.preprocessing.Hashing(num_bins=3)\n inp = [['A'], ['B'], ['C'], ['D'], ['E']]\n layer(inp)\n \u003ctf.Tensor: shape=(5, 1), dtype=int64, numpy=\n array([[1],\n [0],\n [1],\n [1],\n [2]])\u003e\n\nExample (FarmHash64) with list of inputs:\n\u003e \u003e \u003e layer = tf.keras.layers.experimental.preprocessing.Hashing(num_bins=3)\n\u003e \u003e \u003e inp_1 = \\[\\['A'\\], \\['B'\\], \\['C'\\], \\['D'\\], \\['E'\\]\\]\n\u003e \u003e \u003e inp_2 = np.asarray(\\[\\[5\\], \\[4\\], \\[3\\], \\[2\\], \\[1\\]\\])\n\u003e \u003e \u003e layer(\\[inp_1, inp_2\\])\n\nExample (SipHash64): \n\n layer = tf.keras.layers.experimental.preprocessing.Hashing(num_bins=3,\n salt=[133, 137])\n inp = [['A'], ['B'], ['C'], ['D'], ['E']]\n layer(inp)\n \u003ctf.Tensor: shape=(5, 1), dtype=int64, numpy=\n array([[1],\n [2],\n [1],\n [0],\n [2]])\u003e\n\nExample (Siphash64 with a single integer, same as `salt=[133, 133]` \n\n layer = tf.keras.layers.experimental.preprocessing.Hashing(num_bins=3,\n salt=133)\n inp = [['A'], ['B'], ['C'], ['D'], ['E']]\n layer(inp)\n \u003ctf.Tensor: shape=(5, 1), dtype=int64, numpy=\n array([[0],\n [0],\n [2],\n [1],\n [0]])\u003e\n\nReference: [SipHash with salt](https://www.131002.net/siphash/siphash.pdf)\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n| Arguments --------- ||\n|------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|\n| `num_bins` | Number of hash bins. |\n| `salt` | A single unsigned integer or None. If passed, the hash function used will be SipHash64, with these values used as an additional input (known as a \"salt\" in cryptography). These should be non-zero. Defaults to `None` (in that case, the FarmHash64 hash function is used). It also supports tuple/list of 2 unsigned integer numbers, see reference paper for details. |\n| `name` | Name to give to the layer. |\n| `**kwargs` | Keyword arguments to construct a layer. |\n\n\u003cbr /\u003e\n\nInput shape: A single or list of string, int32 or int64 `Tensor`,\n`SparseTensor` or `RaggedTensor` of shape `[batch_size, ...,]`\n\nOutput shape: An int64 `Tensor`, `SparseTensor` or `RaggedTensor` of shape\n`[batch_size, ...]`. If any input is `RaggedTensor` then output is\n`RaggedTensor`, otherwise if any input is `SparseTensor` then output is\n`SparseTensor`, otherwise the output is `Tensor`."]]
View source on GitHub