Package com.linkedin.feathr.common

Class ColumnUtils

java.lang.Object
com.linkedin.feathr.common.ColumnUtils
Direct Known Subclasses:
SimpleWriteableTuple, UniversalTensorBuilder
public class ColumnUtils extends Object
Utility function to build internal flat array representation based on the types of the input.

  • Field Details

    • _columnIndex

      public final int[] _columnIndex

    • _intArity

      public int _intArity

    • _longArity

      public int _longArity

    • _stringArity

      public int _stringArity

    • _floatArity

      public int _floatArity

    • _doubleArity

      public int _doubleArity

    • _booleanArity

      public int _booleanArity

    • _bytesArity

      public int _bytesArity

    • _columnTypes

      public final Representable[] _columnTypes

  • Constructor Details

  • Method Details

    • getColumnIndex

      public int[] getColumnIndex()

    • getIntArity

      public int getIntArity()

    • getLongArity

      public int getLongArity()

    • getStringArity

      public int getStringArity()

    • getFloatArity

      public int getFloatArity()

    • getDoubleArity

      public int getDoubleArity()

    • getBooleanArity

      public int getBooleanArity()

    • getBytesArity

      public int getBytesArity()

    • getColumnTypes

      public Representable[] getColumnTypes()

    • getDimArity

      public static int getDimArity(Representable[] types)
      Compute the number of dimension (non-value) columns This assumes that all value columns are at end and are of type FLOAT or DOUBLE. For other value types, assumes only one value column
      Parameters:
      types - column types
      Returns:
      number of dimension(non-value) columns. For non float/double columns, returns types.length - 1