src/main/java/org/apache/datasketches/pig/hll/DataToSketch.java [88:175]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    super();
    this.lgK_ = lgK;
    this.tgtHllType_ = tgtHllType;
  }

  /**
   * Top-level exec function.
   * This method accepts an input Tuple containing a Bag of one or more inner <b>Datum Tuples</b>
   * and returns a single serialized HllSketch as a DataByteArray.
   *
   * <b>Datum Tuple</b> is a Tuple containing a single field, which can be one of the following
   * (Java type: Pig type):
   * <ul>
   *   <li>Byte: BYTE</li>
   *   <li>Integer: INTEGER</li>
   *   <li>Long: LONG</li>
   *   <li>Float: FLOAT</li>
   *   <li>Double: DOUBLE</li>
   *   <li>String: CHARARRAY</li>
   *   <li>DataByteArray: BYTEARRAY</li>
   * </ul>
   *
   * <p><b>Note</b> Strings as values are normally typed as DataType.CHARARRAY, which will be
   * encoded as UTF-8 prior to being submitted to the sketch. If the user requires a different
   * encoding for cross-platform compatibility, it is recommended that these values be encoded prior
   * to being submitted and then typed as a DataType.BYTEARRAY.</p>
   *
   * @param inputTuple A tuple containing a single bag, containing Datum Tuples.
   * @return serialized HllSketch
   * @see "org.apache.pig.EvalFunc.exec(org.apache.pig.data.Tuple)"
   * @throws IOException from Pig.
   */

  @Override
  public DataByteArray exec(final Tuple inputTuple) throws IOException {
    if (this.isFirstCall_) {
      Logger.getLogger(getClass()).info("Exec was used");
      this.isFirstCall_ = false;
    }
    if (inputTuple == null || inputTuple.size() == 0) {
      if (this.emptySketch_ == null) {
        this.emptySketch_ = 
            new DataByteArray(new HllSketch(this.lgK_, this.tgtHllType_).toCompactByteArray());
      }
      return this.emptySketch_;
    }
    final Union union = new Union(this.lgK_);
    final DataBag bag = (DataBag) inputTuple.get(0);
    updateUnion(bag, union);
    return new DataByteArray(union.getResult(this.tgtHllType_).toCompactByteArray());
  }

  /**
   * An <i>Accumulator</i> version of the standard <i>exec()</i> method. Like <i>exec()</i>,
   * accumulator is called with a bag of Datum Tuples. Unlike <i>exec()</i>, it doesn't serialize the
   * result at the end. Instead, it can be called multiple times, each time with another bag of
   * Datum Tuples to be input to the sketch.
   *
   * @param inputTuple A tuple containing a single bag, containing Datum Tuples.
   * @see #exec
   * @see "org.apache.pig.Accumulator.accumulate(org.apache.pig.data.Tuple)"
   * @throws IOException by Pig
   */
  @Override
  public void accumulate(final Tuple inputTuple) throws IOException {
    if (this.isFirstCall_) {
      Logger.getLogger(getClass()).info("Accumulator was used");
      this.isFirstCall_ = false;
    }
    if (inputTuple == null || inputTuple.size() == 0) { return; }
    final DataBag bag = (DataBag) inputTuple.get(0);
    if (bag == null) { return; }
    if (this.accumUnion_ == null) {
      this.accumUnion_ = new Union(this.lgK_);
    }
    updateUnion(bag, this.accumUnion_);
  }

  /**
   * Returns the sketch that has been built up by multiple calls to {@link #accumulate}.
   *
   * @return serialized HllSketch
   * @see "org.apache.pig.Accumulator.getValue()"
   */
  @Override
  public DataByteArray getValue() {
    if (this.accumUnion_ == null) {
      if (this.emptySketch_ == null) {
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -



src/main/java/org/apache/datasketches/pig/hll/UnionSketch.java [89:161]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    super();
    this.lgK_ = lgK;
    this.tgtHllType_ = tgtHllType;
  }

  /**
   * Top-level exec function.
   * This method accepts an input Tuple containing a Bag of one or more inner <b>Sketch Tuples</b>
   * and returns a single serialized HllSketch as a DataByteArray.
   *
   * <b>Sketch Tuple</b> is a Tuple containing a single DataByteArray (BYTEARRAY in Pig), which
   * is a serialized HllSketch.
   *
   * @param inputTuple A tuple containing a single bag, containing Sketch Tuples.
   * @return serialized HllSketch
   * @see "org.apache.pig.EvalFunc.exec(org.apache.pig.data.Tuple)"
   * @throws IOException from Pig.
   */
  @Override
  public DataByteArray exec(final Tuple inputTuple) throws IOException {
    if (this.isFirstCall_) {
      Logger.getLogger(getClass()).info("Exec was used");
      this.isFirstCall_ = false;
    }
    if (inputTuple == null || inputTuple.size() == 0) {
      if (this.emptySketch_ == null) {
        this.emptySketch_ = 
          new DataByteArray(new HllSketch(this.lgK_, this.tgtHllType_).toCompactByteArray());
      }
      return this.emptySketch_;
    }
    final Union union = new Union(this.lgK_);
    final DataBag bag = (DataBag) inputTuple.get(0);
    updateUnion(bag, union);
    return new DataByteArray(union.getResult(this.tgtHllType_).toCompactByteArray());
  }

  /**
   * An <i>Accumulator</i> version of the standard <i>exec()</i> method. Like <i>exec()</i>,
   * accumulator is called with a bag of Sketch Tuples. Unlike <i>exec()</i>, it doesn't serialize the
   * result at the end. Instead, it can be called multiple times, each time with another bag of
   * Sketch Tuples to be input to the union.
   *
   * @param inputTuple A tuple containing a single bag, containing Sketch Tuples.
   * @see #exec
   * @see "org.apache.pig.Accumulator.accumulate(org.apache.pig.data.Tuple)"
   * @throws IOException by Pig
   */
  @Override
  public void accumulate(final Tuple inputTuple) throws IOException {
    if (this.isFirstCall_) {
      Logger.getLogger(getClass()).info("Accumulator was used");
      this.isFirstCall_ = false;
    }
    if (inputTuple == null || inputTuple.size() == 0) { return; }
    final DataBag bag = (DataBag) inputTuple.get(0);
    if (bag == null) { return; }
    if (this.accumUnion_ == null) {
      this.accumUnion_ = new Union(this.lgK_);
    }
    updateUnion(bag, this.accumUnion_);
  }

  /**
   * Returns the sketch that has been built up by multiple calls to {@link #accumulate}.
   *
   * @return Sketch Tuple. (see {@link #exec} for return tuple format)
   * @see "org.apache.pig.Accumulator.getValue()"
   */
  @Override
  public DataByteArray getValue() {
    if (this.accumUnion_ == null) {
      if (this.emptySketch_ == null) {
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -