UnitVectorLayer

UnitVectorLayer[n]

represents a net layer that transforms integers between 1 and n into n-dimensional unit vectors.

UnitVectorLayer[]

leaves the n to be inferred from context.

Details and Options

  • UnitVectorLayer operates either on integers or arrays of integers, producing a vector or array, respectively.
  • Specifying UnitVectorLayer[n] will produce a net that infers the n when a NetEncoder[…] that produces integers is connected to its input.
  • UnitVectorLayer[…][input] explicitly computes the output from applying the layer to input.
  • UnitVectorLayer[…][{input1,input2,…}] explicitly computes an output for each of the inputi.
  • When given a NumericArray as input, the output will be a NumericArray.
  • UnitVectorLayer is typically used inside NetChain, NetGraph, etc.
  • UnitVectorLayer exposes the following ports for use in NetGraph etc.:
  • "Input"an integer or array of integers
    "Output"an n-dimensional numerical vector or array of vectors
  • UnitVectorLayer[size,"Input"->shape] allows the shape of the input to be specified. Possible forms for shape are:
  • NetEncoder[…]encoder producing an integer or array of integers
    "Integer"a single integer
    da vector of integers of length d
    {d1,d2,…}an array of integers of dimensions d1×d2×…
    "Varying"a vector of integers whose length is variable
    {"Varying",d2,d3,…}an array of integers whose first dimension is variable and remaining dimensions are d2×d3×…
  • Options[UnitVectorLayer] gives the list of default options to construct the layer. Options[UnitVectorLayer[…]] gives the list of default options to evaluate the layer on some data.
  • Information[UnitVectorLayer[…]] gives a report about the layer.
  • Information[UnitVectorLayer[…],prop] gives the value of the property prop of UnitVectorLayer[…]. Possible properties are the same as for NetGraph.

Examples

open allclose all

Basic Examples  (2)

Create a UnitVectorLayer that will accept an array containing 1 or 2 as input and produce an array of "one-hot" vectors:

Create a UnitVectorLayer that will accept 1, 2 or 3 as input and produce a "one-hot" vector:

Apply the layer to an integer to produce an output vector:

Scope  (3)

Ports  (3)

Create a UnitVectorLayer with the input specified to be an integer between 1 and 3:

Apply the layer to an input:

The layer threads across a batch of inputs:

Create a UnitVectorLayer with the input specified to be a sequence of integers between 1 and 3:

Apply the layer to a sequence:

Apply the layer to a batch of sequences of varying lengths:

Explicitly specify that the input is a matrix of integers between 1 and 3:

Apply the input to a matrix to obtain a three-dimensional array:

Applications  (1)

Use a NetEncoder in conjunction with a UnitVectorLayer:

Properties & Relations  (1)

UnitVectorLayer is related to the n-dimensional form of UnitVector:

Apply both to an integer:

Possible Issues  (1)

Out-of-bounds integers will trigger a failure:

Wolfram Research (2017), UnitVectorLayer, Wolfram Language function, https://reference.wolfram.com/language/ref/UnitVectorLayer.html.

Text

Wolfram Research (2017), UnitVectorLayer, Wolfram Language function, https://reference.wolfram.com/language/ref/UnitVectorLayer.html.

CMS

Wolfram Language. 2017. "UnitVectorLayer." Wolfram Language & System Documentation Center. Wolfram Research. https://reference.wolfram.com/language/ref/UnitVectorLayer.html.

APA

Wolfram Language. (2017). UnitVectorLayer. Wolfram Language & System Documentation Center. Retrieved from https://reference.wolfram.com/language/ref/UnitVectorLayer.html

BibTeX

@misc{reference.wolfram_2024_unitvectorlayer, author="Wolfram Research", title="{UnitVectorLayer}", year="2017", howpublished="\url{https://reference.wolfram.com/language/ref/UnitVectorLayer.html}", note=[Accessed: 21-December-2024 ]}

BibLaTeX

@online{reference.wolfram_2024_unitvectorlayer, organization={Wolfram Research}, title={UnitVectorLayer}, year={2017}, url={https://reference.wolfram.com/language/ref/UnitVectorLayer.html}, note=[Accessed: 21-December-2024 ]}