use of org.apache.flink.types.StringValue in project flink by apache.
the class GraphMetrics method main.
public static void main(String[] args) throws Exception {
// Set up the execution environment
final ExecutionEnvironment env = ExecutionEnvironment.getExecutionEnvironment();
env.getConfig().enableObjectReuse();
ParameterTool parameters = ParameterTool.fromArgs(args);
env.getConfig().setGlobalJobParameters(parameters);
if (!parameters.has("directed")) {
throw new ProgramParametrizationException(getUsage("must declare execution mode as '--directed true' or '--directed false'"));
}
boolean directedAlgorithm = parameters.getBoolean("directed");
GraphAnalytic vm;
GraphAnalytic em;
switch(parameters.get("input", "")) {
case "csv":
{
String lineDelimiter = StringEscapeUtils.unescapeJava(parameters.get("input_line_delimiter", CsvOutputFormat.DEFAULT_LINE_DELIMITER));
String fieldDelimiter = StringEscapeUtils.unescapeJava(parameters.get("input_field_delimiter", CsvOutputFormat.DEFAULT_FIELD_DELIMITER));
GraphCsvReader reader = Graph.fromCsvReader(parameters.getRequired("input_filename"), env).ignoreCommentsEdges("#").lineDelimiterEdges(lineDelimiter).fieldDelimiterEdges(fieldDelimiter);
switch(parameters.get("type", "")) {
case "integer":
{
Graph<LongValue, NullValue, NullValue> graph = reader.keyType(LongValue.class);
if (directedAlgorithm) {
if (parameters.getBoolean("simplify", false)) {
graph = graph.run(new org.apache.flink.graph.asm.simple.directed.Simplify<LongValue, NullValue, NullValue>());
}
vm = graph.run(new org.apache.flink.graph.library.metric.directed.VertexMetrics<LongValue, NullValue, NullValue>());
em = graph.run(new org.apache.flink.graph.library.metric.directed.EdgeMetrics<LongValue, NullValue, NullValue>());
} else {
if (parameters.getBoolean("simplify", false)) {
graph = graph.run(new org.apache.flink.graph.asm.simple.undirected.Simplify<LongValue, NullValue, NullValue>(false));
}
vm = graph.run(new org.apache.flink.graph.library.metric.undirected.VertexMetrics<LongValue, NullValue, NullValue>());
em = graph.run(new org.apache.flink.graph.library.metric.undirected.EdgeMetrics<LongValue, NullValue, NullValue>());
}
}
break;
case "string":
{
Graph<StringValue, NullValue, NullValue> graph = reader.keyType(StringValue.class);
if (directedAlgorithm) {
if (parameters.getBoolean("simplify", false)) {
graph = graph.run(new org.apache.flink.graph.asm.simple.directed.Simplify<StringValue, NullValue, NullValue>());
}
vm = graph.run(new org.apache.flink.graph.library.metric.directed.VertexMetrics<StringValue, NullValue, NullValue>());
em = graph.run(new org.apache.flink.graph.library.metric.directed.EdgeMetrics<StringValue, NullValue, NullValue>());
} else {
if (parameters.getBoolean("simplify", false)) {
graph = graph.run(new org.apache.flink.graph.asm.simple.undirected.Simplify<StringValue, NullValue, NullValue>(false));
}
vm = graph.run(new org.apache.flink.graph.library.metric.undirected.VertexMetrics<StringValue, NullValue, NullValue>());
em = graph.run(new org.apache.flink.graph.library.metric.undirected.EdgeMetrics<StringValue, NullValue, NullValue>());
}
}
break;
default:
throw new ProgramParametrizationException(getUsage("invalid CSV type"));
}
}
break;
case "rmat":
{
int scale = parameters.getInt("scale", DEFAULT_SCALE);
int edgeFactor = parameters.getInt("edge_factor", DEFAULT_EDGE_FACTOR);
RandomGenerableFactory<JDKRandomGenerator> rnd = new JDKRandomGeneratorFactory();
long vertexCount = 1L << scale;
long edgeCount = vertexCount * edgeFactor;
Graph<LongValue, NullValue, NullValue> graph = new RMatGraph<>(env, rnd, vertexCount, edgeCount).generate();
if (directedAlgorithm) {
if (scale > 32) {
Graph<LongValue, NullValue, NullValue> newGraph = graph.run(new org.apache.flink.graph.asm.simple.directed.Simplify<LongValue, NullValue, NullValue>());
vm = newGraph.run(new org.apache.flink.graph.library.metric.directed.VertexMetrics<LongValue, NullValue, NullValue>());
em = newGraph.run(new org.apache.flink.graph.library.metric.directed.EdgeMetrics<LongValue, NullValue, NullValue>());
} else {
Graph<IntValue, NullValue, NullValue> newGraph = graph.run(new TranslateGraphIds<LongValue, IntValue, NullValue, NullValue>(new LongValueToUnsignedIntValue())).run(new org.apache.flink.graph.asm.simple.directed.Simplify<IntValue, NullValue, NullValue>());
vm = newGraph.run(new org.apache.flink.graph.library.metric.directed.VertexMetrics<IntValue, NullValue, NullValue>());
em = newGraph.run(new org.apache.flink.graph.library.metric.directed.EdgeMetrics<IntValue, NullValue, NullValue>());
}
} else {
boolean clipAndFlip = parameters.getBoolean("clip_and_flip", DEFAULT_CLIP_AND_FLIP);
if (scale > 32) {
Graph<LongValue, NullValue, NullValue> newGraph = graph.run(new org.apache.flink.graph.asm.simple.undirected.Simplify<LongValue, NullValue, NullValue>(clipAndFlip));
vm = newGraph.run(new org.apache.flink.graph.library.metric.undirected.VertexMetrics<LongValue, NullValue, NullValue>());
em = newGraph.run(new org.apache.flink.graph.library.metric.undirected.EdgeMetrics<LongValue, NullValue, NullValue>());
} else {
Graph<IntValue, NullValue, NullValue> newGraph = graph.run(new TranslateGraphIds<LongValue, IntValue, NullValue, NullValue>(new LongValueToUnsignedIntValue())).run(new org.apache.flink.graph.asm.simple.undirected.Simplify<IntValue, NullValue, NullValue>(clipAndFlip));
vm = newGraph.run(new org.apache.flink.graph.library.metric.undirected.VertexMetrics<IntValue, NullValue, NullValue>());
em = newGraph.run(new org.apache.flink.graph.library.metric.undirected.EdgeMetrics<IntValue, NullValue, NullValue>());
}
}
}
break;
default:
throw new ProgramParametrizationException(getUsage("invalid input type"));
}
env.execute("Graph Metrics");
System.out.println();
System.out.print("Vertex metrics:\n ");
System.out.println(vm.getResult().toString().replace(";", "\n "));
System.out.println();
System.out.print("Edge metrics:\n ");
System.out.println(em.getResult().toString().replace(";", "\n "));
JobExecutionResult result = env.getLastJobExecutionResult();
NumberFormat nf = NumberFormat.getInstance();
System.out.println();
System.out.println("Execution runtime: " + nf.format(result.getNetRuntime()) + " ms");
}
use of org.apache.flink.types.StringValue in project flink by apache.
the class SimpleStringUtilsTest method testReplaceNonWordChars.
@Test
public void testReplaceNonWordChars() {
StringValue testString = new StringValue("TEST123_@");
SimpleStringUtils.replaceNonWordChars(testString, '!');
assertEquals(new StringValue("TEST123_!"), testString);
}
use of org.apache.flink.types.StringValue in project flink by apache.
the class SimpleStringUtilsTest method testTokenizerOnStringWithNexToken.
@Test
public void testTokenizerOnStringWithNexToken() {
StringValue testString = new StringValue("test test");
SimpleStringUtils.WhitespaceTokenizer tokenizer = new WhitespaceTokenizer();
tokenizer.setStringToTokenize(testString);
assertTrue(tokenizer.next(testString));
}
use of org.apache.flink.types.StringValue in project flink by apache.
the class SimpleStringUtilsTest method testTokenizerOnStringWithoutNexToken.
@Test
public void testTokenizerOnStringWithoutNexToken() {
StringValue testString = new StringValue("test");
SimpleStringUtils.WhitespaceTokenizer tokenizer = new WhitespaceTokenizer();
tokenizer.setStringToTokenize(testString);
//first token
tokenizer.next(testString);
//next token is not exist
assertFalse(tokenizer.next(testString));
}
use of org.apache.flink.types.StringValue in project flink by apache.
the class SimpleStringUtilsTest method testTokenizerOnStringOnlyWithDelimiter.
@Test
public void testTokenizerOnStringOnlyWithDelimiter() {
StringValue testString = new StringValue(" ");
SimpleStringUtils.WhitespaceTokenizer tokenizer = new WhitespaceTokenizer();
tokenizer.setStringToTokenize(testString);
assertFalse(tokenizer.next(testString));
}
Aggregations