enrico.ottonello
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

merged with master

This commit is contained in:
Miriam Baglioni 2020-05-07 11:48:21 +02:00
parent e124278934 5111671e62
commit fb405275f7
1 changed files with 8 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
dhp-workflows/dhp-propagation/src/main/java/eu/dnetlib/dhp/countrypropagation/SparkCountryPropagationJob3.java
View File

@ -71,20 +71,15 @@ public class SparkCountryPropagationJob3 {
Class<? extends Result> resultClazz = (Class<? extends Result>) Class.forName(resultClassName); Class<? extends Result> resultClazz = (Class<? extends Result>) Class.forName(resultClassName);
SparkConf conf = new SparkConf(); SparkConf conf = new SparkConf();
conf.registerKryoClasses(ModelSupport.getOafModelClasses());
runWithSparkSession( runWithSparkSession(
conf, conf,
isSparkSessionManaged, isSparkSessionManaged,
spark -> { spark -> execPropagation(
spark,
execPropagation( inputPath,
spark, outputPath,
inputPath, resultClazz,
outputPath, saveGraph));
resultClazz,
saveGraph);
});
} }
private static <R extends Result> void execPropagation( private static <R extends Result> void execPropagation(
@ -101,8 +96,8 @@ public class SparkCountryPropagationJob3 {
spark spark
.read() .read()
.json(inputPath) .json(inputPath)
.as(Encoders.kryo(resultClazz)) .as(Encoders.bean(resultClazz))
.groupByKey((MapFunction<R, String>) result1 -> result1.getId(), Encoders.STRING()) .groupByKey((MapFunction<R, String>) r -> r.getId(), Encoders.STRING())
.mapGroups(getCountryMergeFn(resultClazz), Encoders.bean(resultClazz)) .mapGroups(getCountryMergeFn(resultClazz), Encoders.bean(resultClazz))
.write() .write()
.option("compression", "gzip") .option("compression", "gzip")