Package deepnetts.data

Class TabularDataSet<E extends MLDataItem>

  • Type Parameters:
    E - Type of elements in this data set.
    All Implemented Interfaces:
    java.lang.Iterable<E>, javax.visrec.ml.data.DataSet<E>
    Direct Known Subclasses:
    ImageSet
    public class TabularDataSet<E extends MLDataItem>
extends javax.visrec.ml.data.BasicDataSet<E>
    Basic data set used for training neural networks in deep netts.

    Note: implements DataSet from visrec api, and specify data set elements. Extends BasicDataSet from visrec.ml layer

    Author:
    Zoran Sevarac

    • Nested Class Summary

      Nested Classes 
      Modifier and Type Class Description
      static class  TabularDataSet.Item
      Represents a basic data set item (single row) with input tensor and target vector in a data set.

    • Field Summary

      Fields 
      Modifier and Type Field Description
      protected java.lang.String[] columnNames  

      • Fields inherited from class javax.visrec.ml.data.BasicDataSet

        items

    • Constructor Summary

      Constructors 
      Modifier Constructor Description
      protected TabularDataSet()  
        TabularDataSet​(int numInputs, int numOutputs)
      Create a new instance of BasicDataSet with specified size of input and output.

    • Method Summary

      Modifier and Type Method Description
      java.lang.String[] getColumnNames()  
      int getNumInputs()  
      int getNumOutputs()  
      void setColumnNames​(java.lang.String[] columnNames)  
      void shuffle()
      Shuffles the data set items using the default random generator.
      void shuffle​(int seed)
      Shuffles data set items using java random generator initializes with specified seed
      javax.visrec.ml.data.DataSet[] split​(double... parts)
      Splits data set into several parts specified by the input parameter partSizes.
      javax.visrec.ml.data.DataSet[] split​(int parts)
      Split data set into specified number of part of equal sizes.

      • Methods inherited from class javax.visrec.ml.data.BasicDataSet

        getColumns, getItems, getTargetColumnsNames, setAsTargetColumns, setAsTargetColumns, setColumns

      • Methods inherited from class java.lang.Object

        clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

      • Methods inherited from interface javax.visrec.ml.data.DataSet

        add, addAll, clear, get, isEmpty, iterator, shuffle, size, split, split, split, stream

      • Methods inherited from interface java.lang.Iterable

        forEach, spliterator

    • Field Detail

      • columnNames

        protected java.lang.String[] columnNames

    • Constructor Detail

      • TabularDataSet

        public TabularDataSet​(int numInputs,
                      int numOutputs)
        Create a new instance of BasicDataSet with specified size of input and output.
        Parameters:
        numInputs - number of input features
        numOutputs - number of output features

    • Method Detail

      • split

        public javax.visrec.ml.data.DataSet[] split​(int parts)
        Split data set into specified number of part of equal sizes. Utility method used during cross-validation Note: this could be default method
        Parameters:
        parts -
        Returns:

      • split

        public javax.visrec.ml.data.DataSet[] split​(double... parts)
        Splits data set into several parts specified by the input parameter partSizes. Values of partSizes parameter represent the sizes of data set parts that will be returned. Part sizes are decimal values that represent percents, cannot be negative or zero, and their sum must be 1
        Specified by:
        split in interface javax.visrec.ml.data.DataSet<E extends MLDataItem>
        Overrides:
        split in class javax.visrec.ml.data.BasicDataSet<E extends MLDataItem>
        Parameters:
        parts - sizes of the parts in percents
        Returns:
        parts of the data set of specified size

      • shuffle

        public void shuffle()
        Shuffles the data set items using the default random generator. Default rng can be initialized independently

      • shuffle

        public void shuffle​(int seed)
        Shuffles data set items using java random generator initializes with specified seed
        Parameters:
        seed - a seed number to initialize random generator
        See Also:
        Random

      • getColumnNames

        public java.lang.String[] getColumnNames()
        Overrides:
        getColumnNames in class javax.visrec.ml.data.BasicDataSet<E extends MLDataItem>

      • setColumnNames

        public void setColumnNames​(java.lang.String[] columnNames)
        Overrides:
        setColumnNames in class javax.visrec.ml.data.BasicDataSet<E extends MLDataItem>