flink-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From rmetz...@apache.org
Subject git commit: [FLINK-828] Implement accumulator example
Date Tue, 15 Jul 2014 18:18:50 GMT
Repository: incubator-flink
Updated Branches:
  refs/heads/master 3002258f8 -> 5d4590ae3


[FLINK-828] Implement accumulator example

- shows how to build a custom accumulator (for vectors)
- employ that accumulator for obtaining filter statistics

This closes #55.


Project: http://git-wip-us.apache.org/repos/asf/incubator-flink/repo
Commit: http://git-wip-us.apache.org/repos/asf/incubator-flink/commit/5d4590ae
Tree: http://git-wip-us.apache.org/repos/asf/incubator-flink/tree/5d4590ae
Diff: http://git-wip-us.apache.org/repos/asf/incubator-flink/diff/5d4590ae

Branch: refs/heads/master
Commit: 5d4590ae38f3120c66e80fe354bbf4f29ed6fb5b
Parents: 3002258
Author: Sebastian Kruse <sebastian.kruse@hpi.de>
Authored: Tue Jul 1 15:22:18 2014 +0200
Committer: Robert Metzger <rmetzger@apache.org>
Committed: Mon Jul 14 15:38:53 2014 +0200

----------------------------------------------------------------------
 .../relational/EmptyFieldsCountAccumulator.java | 261 +++++++++++++++++++
 1 file changed, 261 insertions(+)
----------------------------------------------------------------------


http://git-wip-us.apache.org/repos/asf/incubator-flink/blob/5d4590ae/flink-examples/flink-java-examples/src/main/java/org/apache/flink/example/java/relational/EmptyFieldsCountAccumulator.java
----------------------------------------------------------------------
diff --git a/flink-examples/flink-java-examples/src/main/java/org/apache/flink/example/java/relational/EmptyFieldsCountAccumulator.java
b/flink-examples/flink-java-examples/src/main/java/org/apache/flink/example/java/relational/EmptyFieldsCountAccumulator.java
new file mode 100644
index 0000000..61b90dd
--- /dev/null
+++ b/flink-examples/flink-java-examples/src/main/java/org/apache/flink/example/java/relational/EmptyFieldsCountAccumulator.java
@@ -0,0 +1,261 @@
+/**
+ * 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.flink.example.java.relational;
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.List;
+
+import org.apache.flink.api.common.JobExecutionResult;
+import org.apache.flink.api.common.accumulators.Accumulator;
+import org.apache.flink.api.java.DataSet;
+import org.apache.flink.api.java.ExecutionEnvironment;
+import org.apache.flink.api.java.functions.FilterFunction;
+import org.apache.flink.api.java.tuple.Tuple;
+import org.apache.flink.api.java.tuple.Tuple3;
+import org.apache.flink.configuration.Configuration;
+import org.apache.flink.core.memory.DataInputView;
+import org.apache.flink.core.memory.DataOutputView;
+
+
+/**
+ * This program filters lines from a CSV file with empty fields. In doing so, it counts the
number of empty fields per
+ * column within a CSV file using a custom accumulator for vectors. In this context, empty
fields are those, that at
+ * most contain whitespace characters like space and tab.
+ * <p>
+ * The input file is a plain text CSV file with the semicolon as field separator and double
quotes as field delimiters
+ * and three columns. See {@link #getDataSet(ExecutionEnvironment)} for configuration.
+ * <p>
+ * Usage: <code>FilterAndCountIncompleteLines [&lt;input file path&gt; [&lt;result
path&gt;]]</code> <br>
+ * <p>
+ * This example shows how to use:
+ * <ul>
+ * <li>custom accumulators
+ * <li>tuple data types
+ * <li>inline-defined functions
+ * </ul>
+ */
+@SuppressWarnings("serial")
+public class EmptyFieldsCountAccumulator {
+
+	// *************************************************************************
+	// PROGRAM
+	// *************************************************************************
+
+	private static final String EMPTY_FIELD_ACCUMULATOR = "empty-fields";
+
+	public static void main(final String[] args) throws Exception {
+
+		if (!parseParameters(args)) {
+			return;
+		}
+
+		final ExecutionEnvironment env = ExecutionEnvironment.getExecutionEnvironment();
+
+		// get the data set
+		final DataSet<Tuple> file = getDataSet(env);
+
+		// filter lines with empty fields
+		final DataSet<Tuple> filteredLines = file.filter(new EmptyFieldFilter());
+
+		// Here, we could do further processing with the filtered lines...
+		
+		// output the filtered lines
+		if (outputPath == null) {
+			filteredLines.print();
+		} else {
+			filteredLines.writeAsCsv(outputPath);
+		}
+
+		// execute program
+		final JobExecutionResult result = env.execute("Accumulator example");
+
+		// get the accumulator result via its registration key
+		final List<Integer> emptyFields = result.getAccumulatorResult(EMPTY_FIELD_ACCUMULATOR);
+		System.out.format("Number of detected empty fields per column: %s\n", emptyFields);
+
+	}
+
+	// *************************************************************************
+	// UTIL METHODS
+	// *************************************************************************
+
+	private static String filePath;
+	private static String outputPath;
+
+	private static boolean parseParameters(final String[] programArguments) {
+
+		if (programArguments.length >= 3) {
+			System.err.println("Usage: FilterAndCountIncompleteLines [<input file path> [<result
path>]]");
+			return false;
+		}
+
+		if (programArguments.length >= 1) {
+			filePath = programArguments[0];
+			if (programArguments.length == 2) {
+				outputPath = programArguments[1];
+			}
+		}
+
+		return true;
+	}
+
+	@SuppressWarnings("unchecked")
+	private static DataSet<Tuple> getDataSet(final ExecutionEnvironment env) {
+
+		DataSet<? extends Tuple> source;
+		if (filePath == null) {
+			source = env.fromCollection(getExampleInputTuples());
+
+		} else {
+			source = env
+					.readCsvFile(filePath)
+					.fieldDelimiter(';')
+					.types(String.class, String.class, String.class);
+
+		}
+
+		return (DataSet<Tuple>) source;
+	}
+
+	private static Collection<Tuple3<String, String, String>> getExampleInputTuples()
{
+		Collection<Tuple3<String, String, String>> inputTuples = new ArrayList<Tuple3<String,
String, String>>();
+		inputTuples.add(new Tuple3<String, String, String>("John", "Doe", "Foo Str."));
+		inputTuples.add(new Tuple3<String, String, String>("Joe", "Johnson", ""));
+		inputTuples.add(new Tuple3<String, String, String>(null, "Kate Morn", "Bar Blvd."));
+		inputTuples.add(new Tuple3<String, String, String>("Tim", "Rinny", ""));
+		inputTuples.add(new Tuple3<String, String, String>("Alicia", "Jackson", "  "));
+		return inputTuples;
+	}
+
+	/**
+	 * This function filters all incoming tuples that have one or more empty fields.
+	 * In doing so, it also counts the number of empty fields per attribute with an accumulator
(registered under 
+	 * {@link EmptyFieldsCountAccumulator#EMPTY_FIELD_ACCUMULATOR}).
+	 */
+	public static final class EmptyFieldFilter extends FilterFunction<Tuple> {
+
+		// create a new accumulator in each filter function instance
+		// accumulators can be merged later on
+		private final VectorAccumulator emptyFieldCounter = new VectorAccumulator();
+
+		@Override
+		public void open(final Configuration parameters) throws Exception {
+			super.open(parameters);
+
+			// register the accumulator instance
+			getRuntimeContext().addAccumulator(EMPTY_FIELD_ACCUMULATOR,
+					this.emptyFieldCounter);
+		}
+
+		@Override
+		public boolean filter(final Tuple t) {
+			boolean containsEmptyFields = false;
+
+			// iterate over the tuple fields looking for empty ones
+			for (int pos = 0; pos < t.getArity(); pos++) {
+
+				final String field = t.getField(pos);
+				if (field == null || field.trim().isEmpty()) {
+					containsEmptyFields = true;
+
+					// if an empty field is encountered, update the
+					// accumulator
+					this.emptyFieldCounter.add(pos);
+				}
+			}
+
+			return !containsEmptyFields;
+		}
+	}
+
+	/**
+	 * This accumulator lets you increase vector components distributedly. The {@link #add(Integer)}
method lets you
+	 * increase the <i>n</i>-th vector component by 1, whereat <i>n</i>
is the methods parameter. The size of the vector
+	 * is automatically managed.
+	 */
+	public static class VectorAccumulator implements Accumulator<Integer, List<Integer>>
{
+
+		/** Stores the accumulated vector components. */
+		private final List<Integer> resultVector = new ArrayList<Integer>();
+
+		/**
+		 * Increases the result vector component at the specified position by 1.
+		 */
+		@Override
+		public void add(final Integer position) {
+			updateResultVector(position, 1);
+		}
+
+		/**
+		 * Increases the result vector component at the specified position by the specified delta.
+		 */
+		private void updateResultVector(final int position, final int delta) {
+			// inflate the vector to contain the given position
+			while (this.resultVector.size() <= position) {
+				this.resultVector.add(0);
+			}
+
+			// increment the component value
+			final int component = this.resultVector.get(position);
+			this.resultVector.set(position, component + delta);
+		}
+
+		@Override
+		public List<Integer> getLocalValue() {
+			return this.resultVector;
+		}
+
+		@Override
+		public void resetLocal() {
+			// clear the result vector if the accumulator instance shall be reused
+			this.resultVector.clear();
+		}
+
+		@Override
+		public void merge(final Accumulator<Integer, List<Integer>> other) {
+			// merge two vector accumulators by adding their up their vector components
+			final List<Integer> otherVector = other.getLocalValue();
+			for (int index = 0; index < otherVector.size(); index++) {
+				updateResultVector(index, otherVector.get(index));
+			}
+		}
+
+		@Override
+		public void write(final DataOutputView out) throws IOException {
+			// binary serialization of the result vector:
+			// [number of components, component 0, component 1, ...]
+			out.writeInt(this.resultVector.size());
+			for (final Integer component : this.resultVector) {
+				out.writeInt(component);
+			}
+		}
+
+		@Override
+		public void read(final DataInputView in) throws IOException {
+			// binary deserialization of the result vector
+			final int size = in.readInt();
+			for (int numReadComponents = 0; numReadComponents < size; numReadComponents++) {
+				final int component = in.readInt();
+				this.resultVector.add(component);
+			}
+		}
+
+	}
+}


Mime
View raw message