You are viewing a plain text version of this content. The canonical link for it is here.

Posted to commits@pinot.apache.org by "cbalci (via GitHub)" <gi...@apache.org> on 2023/07/13 16:24:22 UTC

[GitHub] [pinot] cbalci opened a new pull request, #11098: Introduce FrequentStringsSketch and FrequentLongsSketch aggregation functions

cbalci opened a new pull request, #11098:
URL: https://github.com/apache/pinot/pull/11098

   Introducing two approximate aggregation function `FrequentStringsSketch` and `FrequentLongsSketch` for estimating the frequencies of items a dataset in a memory efficient way. Functions are based on the [Apache Datasketches](https://datasketches.apache.org/docs/Frequency/FrequencySketchesOverview.html) library.
   
   **Signature:**
   ```
   FREQUENTLONGSSKETCH(col, maxMapSize=256) -> Base64 encoded sketch object
   FREQUENTSTRINGSSKETCH(col, maxMapSize=256) -> Base64 encoded sketch object
   ```
   
   **Example usage:**
   ```
   select FREQUENTSTRINGSSKETCH(AirlineID, 16) from airlineStats
   ```
   
   | frequentstringssketch(AirlineID)      |
   | ----------- |
   | BAEKAwMAAAADAAAAAA...      |
   
   Which can be used, for example in Java as:
   ```java
   byte[] byteArr = Base64.getDecoder().decode(encodedSketch);
   ItemsSketch<String> sketch = ItemsSketch.getInstance(Memory.wrap(byteArr), new ArrayOfStringsSerDe());
   
   ItemsSketch.Row[] items = sketch.getFrequentItems(ErrorType.NO_FALSE_NEGATIVES);
   for (int i = 0; i < items.length; i++) {
     ItemsSketch.Row item = items[i];
     System.out.printf("Airline: %s, Frequency: %d %n", item.getItem(), item.getEstimate());
   }
   ```
   
   **Testing:**
   Basic aggregation and group by query tests are included in the PR.
   
   **Design:**
   This is a part of a larger effort to improve Datasketches support for Pinot as discussed in [this document](https://docs.google.com/document/d/1ctmKVRi67lpO6x1RYKDvDYf05EZx2Vbs2OnUudYP-bU/edit#heading=h.nctch2wugvub). Feel free to add design related comments on the document as well.
   
   
   `feature` `performance`


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@pinot.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@pinot.apache.org
For additional commands, e-mail: commits-help@pinot.apache.org

[GitHub] [pinot] chenboat commented on a diff in pull request #11098: Introduce FrequentStringsSketch and FrequentLongsSketch aggregation functions

Posted by "chenboat (via GitHub)" <gi...@apache.org>.

chenboat commented on code in PR #11098:
URL: https://github.com/apache/pinot/pull/11098#discussion_r1281307520


##########
pinot-core/src/main/java/org/apache/pinot/core/query/aggregation/function/FrequentLongsSketchAggregationFunction.java:
##########
@@ -0,0 +1,251 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements.  See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership.  The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License.  You may obtain a copy of the License at
+ *
+ *   http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing,
+ * software distributed under the License is distributed on an
+ * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+ * KIND, either express or implied.  See the License for the
+ * specific language governing permissions and limitations
+ * under the License.
+ */
+
+package org.apache.pinot.core.query.aggregation.function;
+
+import com.google.common.base.Preconditions;
+import java.util.List;
+import java.util.Map;
+import org.apache.datasketches.frequencies.LongsSketch;
+import org.apache.datasketches.memory.Memory;
+import org.apache.pinot.common.request.context.ExpressionContext;
+import org.apache.pinot.common.utils.DataSchema;
+import org.apache.pinot.core.common.BlockValSet;
+import org.apache.pinot.core.query.aggregation.AggregationResultHolder;
+import org.apache.pinot.core.query.aggregation.ObjectAggregationResultHolder;
+import org.apache.pinot.core.query.aggregation.groupby.GroupByResultHolder;
+import org.apache.pinot.core.query.aggregation.groupby.ObjectGroupByResultHolder;
+import org.apache.pinot.segment.local.customobject.SerializedFrequentLongsSketch;
+import org.apache.pinot.segment.spi.AggregationFunctionType;
+import org.apache.pinot.spi.data.FieldSpec;
+
+
+/**
+ * <p>
+ *  {@code FrequentLongsSketchAggregationFunction} provides an approximate FrequentItems aggregation function based on
+ *  <a href="https://datasketches.apache.org/docs/Frequency/FrequentItemsOverview.html">Apache DataSketches library</a>.
+ *  It is memory efficient compared to exact counting.
+ * </p>
+ * <p>
+ *   The function takes an INT or LONG column as input and returns a Base64 encoded sketch object which can be
+ *   deserialized and used to estimate the frequency of items in the dataset (how many times they appear).
+ * </p>
+ * <p><b>FREQUENT_STRINGS_SKETCH(col, maxMapSize=256)</b></p>
+ * <p>E.g.:</p>
+ * <ul>
+ *   <li><b>FREQUENT_LONGS_SKETCH(col)</b></li>
+ *   <li><b>FREQUENT_LONGS_SKETCH(col, 1024)</b></li>
+ * </ul>
+ *
+ * <p>
+ *   If the column type is BYTES, the aggregation function will assume it is a serialized FrequentItems data sketch
+ *   of type `LongsSketch`and will attempt to deserialize it for merging with other sketch objects.
+ * </p>
+ *
+ * <p>
+ *   There is a variation of the function (<b>FREQUENT_STRINGS_SKETCH</b>) which accepts STRING type input columns.
+ * </p>
+ */
+public class FrequentLongsSketchAggregationFunction
+    extends BaseSingleInputAggregationFunction<LongsSketch, Comparable<?>> {
+  protected static final int DEFAULT_MAX_MAP_SIZE = 256;
+
+  protected int _maxMapSize;
+
+  public FrequentLongsSketchAggregationFunction(List<ExpressionContext> arguments) {
+    super(arguments.get(0));
+    int numArguments = arguments.size();
+    Preconditions.checkArgument(numArguments == 1 || numArguments == 2,
+        "Expecting 1 or 2 arguments for FrequentLongsSketch function: FREQUENTITEMSSKETCH(column, maxMapSize");
+    _maxMapSize = numArguments == 2 ? arguments.get(1).getLiteral().getIntValue() : DEFAULT_MAX_MAP_SIZE;
+  }
+
+  @Override
+  public AggregationFunctionType getType() {
+    return AggregationFunctionType.FREQUENTLONGSSKETCH;
+  }
+
+  @Override
+  public AggregationResultHolder createAggregationResultHolder() {
+    return new ObjectAggregationResultHolder();
+  }
+
+  @Override
+  public GroupByResultHolder createGroupByResultHolder(int initialCapacity, int maxCapacity) {
+    return new ObjectGroupByResultHolder(initialCapacity, maxCapacity);
+  }
+
+  @Override
+  public void aggregate(int length, AggregationResultHolder aggregationResultHolder,
+      Map<ExpressionContext, BlockValSet> blockValSetMap) {
+    BlockValSet valueSet = blockValSetMap.get(_expression);
+    FieldSpec.DataType valueType = valueSet.getValueType();
+
+    LongsSketch sketch = getOrCreateSketch(aggregationResultHolder);
+
+    if (valueType == FieldSpec.DataType.BYTES) {

Review Comment:
   NIT: can we use switch-case here?



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@pinot.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@pinot.apache.org
For additional commands, e-mail: commits-help@pinot.apache.org