Package org.apache.commons.math3.ml.neuralnet.twod

Class NeuronSquareMesh2D

    • Field Detail

      • serialVersionUID

        private static final long serialVersionUID
        Serial version ID
        See Also:
        Constant Field Values

      • network

        private final Network network
        Underlying network.

      • numberOfRows

        private final int numberOfRows
        Number of rows.

      • numberOfColumns

        private final int numberOfColumns
        Number of columns.

      • wrapRows

        private final boolean wrapRows
        Wrap.

      • wrapColumns

        private final boolean wrapColumns
        Wrap.

      • identifiers

        private final long[][] identifiers
        Mapping of the 2D coordinates (in the rectangular mesh) to the neuron identifiers (attributed by the network instance).

    • Constructor Detail

      • NeuronSquareMesh2D

        NeuronSquareMesh2D​(boolean wrapRowDim,
                   boolean wrapColDim,
                   SquareNeighbourhood neighbourhoodType,
                   double[][][] featuresList)
        Constructor with restricted access, solely used for deserialization.
        Parameters:
        wrapRowDim - Whether to wrap the first dimension (i.e the first and last neurons will be linked together).
        wrapColDim - Whether to wrap the second dimension (i.e the first and last neurons will be linked together).
        neighbourhoodType - Neighbourhood type.
        featuresList - Arrays that will initialize the features sets of the network's neurons.
        Throws:
        NumberIsTooSmallException - if numRows < 2 or numCols < 2.

      • NeuronSquareMesh2D

        public NeuronSquareMesh2D​(int numRows,
                          boolean wrapRowDim,
                          int numCols,
                          boolean wrapColDim,
                          SquareNeighbourhood neighbourhoodType,
                          FeatureInitializer[] featureInit)
        Creates a two-dimensional network composed of square cells: Each neuron not located on the border of the mesh has four neurons linked to it.
        The links are bi-directional.
        The topology of the network can also be a cylinder (if one of the dimensions is wrapped) or a torus (if both dimensions are wrapped).
        Parameters:
        numRows - Number of neurons in the first dimension.
        wrapRowDim - Whether to wrap the first dimension (i.e the first and last neurons will be linked together).
        numCols - Number of neurons in the second dimension.
        wrapColDim - Whether to wrap the second dimension (i.e the first and last neurons will be linked together).
        neighbourhoodType - Neighbourhood type.
        featureInit - Array of functions that will initialize the corresponding element of the features set of each newly created neuron. In particular, the size of this array defines the size of feature set.
        Throws:
        NumberIsTooSmallException - if numRows < 2 or numCols < 2.

      • NeuronSquareMesh2D

        private NeuronSquareMesh2D​(boolean wrapRowDim,
                           boolean wrapColDim,
                           SquareNeighbourhood neighbourhoodType,
                           Network net,
                           long[][] idGrid)
        Constructor with restricted access, solely used for making a deep copy.
        Parameters:
        wrapRowDim - Whether to wrap the first dimension (i.e the first and last neurons will be linked together).
        wrapColDim - Whether to wrap the second dimension (i.e the first and last neurons will be linked together).
        neighbourhoodType - Neighbourhood type.
        net - Underlying network.
        idGrid - Neuron identifiers.

    • Method Detail

      • copy

        public NeuronSquareMesh2D copy()
        Performs a deep copy of this instance. Upon return, the copied and original instances will be independent: Updating one will not affect the other.
        Returns:
        a new instance with the same state as this instance.
        Since:
        3.6

      • iterator

        public java.util.Iterator<Neuron> iterator()
        Specified by:
        iterator in interface java.lang.Iterable<Neuron>
        Since:
        3.6

      • getNetwork

        public Network getNetwork()
        Retrieves the underlying network. A reference is returned (enabling, for example, the network to be trained). This also implies that calling methods that modify the Network topology may cause this class to become inconsistent.
        Returns:
        the network.

      • getNumberOfRows

        public int getNumberOfRows()
        Gets the number of neurons in each row of this map.
        Returns:
        the number of rows.

      • getNumberOfColumns

        public int getNumberOfColumns()
        Gets the number of neurons in each column of this map.
        Returns:
        the number of column.

      • getNeuron

        public Neuron getNeuron​(int row,
                        int col,
                        NeuronSquareMesh2D.HorizontalDirection alongRowDir,
                        NeuronSquareMesh2D.VerticalDirection alongColDir)
        Retrieves the neuron at (location[0], location[1]) in the map. The neuron at position (0, 0) is located at the upper-left corner of the map.
        Parameters:
        row - Row index.
        col - Column index.
        alongRowDir - Direction along the given row (i.e. an offset will be added to the given column index.
        alongColDir - Direction along the given col (i.e. an offset will be added to the given row index.
        Returns:
        the neuron at the requested location, or null if the location is not on the map.
        See Also:
        getNeuron(int,int)

      • getLocation

        private int[] getLocation​(int row,
                          int col,
                          NeuronSquareMesh2D.HorizontalDirection alongRowDir,
                          NeuronSquareMesh2D.VerticalDirection alongColDir)
        Computes the location of a neighbouring neuron. It will return null if the resulting location is not part of the map. Position (0, 0) is at the upper-left corner of the map.
        Parameters:
        row - Row index.
        col - Column index.
        alongRowDir - Direction along the given row (i.e. an offset will be added to the given column index.
        alongColDir - Direction along the given col (i.e. an offset will be added to the given row index.
        Returns:
        an array of length 2 containing the indices of the requested location, or null if that location is not part of the map.
        See Also:
        getNeuron(int,int)

      • createLinks

        private void createLinks()
        Creates the neighbour relationships between neurons.

      • readObject

        private void readObject​(java.io.ObjectInputStream in)
        Prevents proxy bypass.
        Parameters:
        in - Input stream.

      • writeReplace

        private java.lang.Object writeReplace()
        Custom serialization.
        Returns:
        the proxy instance that will be actually serialized.