sabeel
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
0
Fork 0
Code Pull Requests Activity

changed the bl because of changed of the examples for the re implementation of the propagation step

This commit is contained in:
Miriam Baglioni 2020-04-27 10:58:41 +02:00
parent 6135096ef1
commit ec7f166690
1 changed files with 161 additions and 84 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

227
dhp-workflows/dhp-propagation/src/test/java/eu/dnetlib/dhp/projecttoresult/ProjectPropagationJobTest.java
View File

@ -2,6 +2,9 @@ package eu.dnetlib.dhp.projecttoresult;
import com.fasterxml.jackson.databind.ObjectMapper; import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.dhp.schema.oaf.Relation; import eu.dnetlib.dhp.schema.oaf.Relation;
import java.io.IOException;
import java.nio.file.Files;
import java.nio.file.Path;
import org.apache.commons.io.FileUtils; import org.apache.commons.io.FileUtils;
import org.apache.spark.SparkConf; import org.apache.spark.SparkConf;
import org.apache.spark.api.java.JavaRDD; import org.apache.spark.api.java.JavaRDD;
@ -15,9 +18,6 @@ import org.junit.jupiter.api.BeforeAll;
import org.junit.jupiter.api.Test; import org.junit.jupiter.api.Test;
import org.slf4j.Logger; import org.slf4j.Logger;
import org.slf4j.LoggerFactory; import org.slf4j.LoggerFactory;
import java.io.IOException;
import java.nio.file.Files;
import java.nio.file.Path;
public class ProjectPropagationJobTest { public class ProjectPropagationJobTest {
@ -31,7 +31,6 @@ public class ProjectPropagationJobTest {
private static Path workingDir; private static Path workingDir;
@BeforeAll @BeforeAll
public static void beforeAll() throws IOException { public static void beforeAll() throws IOException {
workingDir = Files.createTempDirectory(ProjectPropagationJobTest.class.getSimpleName()); workingDir = Files.createTempDirectory(ProjectPropagationJobTest.class.getSimpleName());
@ -47,13 +46,11 @@ public class ProjectPropagationJobTest {
conf.set("spark.sql.warehouse.dir", workingDir.toString()); conf.set("spark.sql.warehouse.dir", workingDir.toString());
conf.set("hive.metastore.warehouse.dir", workingDir.resolve("warehouse").toString()); conf.set("hive.metastore.warehouse.dir", workingDir.resolve("warehouse").toString());
spark =
spark = SparkSession SparkSession.builder()
.builder()
.appName(ProjectPropagationJobTest.class.getSimpleName()) .appName(ProjectPropagationJobTest.class.getSimpleName())
.config(conf) .config(conf)
.getOrCreate(); .getOrCreate();
} }
@AfterAll @AfterAll
@ -63,123 +60,203 @@ public class ProjectPropagationJobTest {
} }
/** /**
* There are no new relations to be added. All the possible relations have already been linked with the project in the graph * There are no new relations to be added. All the possible relations have already been linked
* with the project in the graph
*
* @throws Exception * @throws Exception
*/ */
@Test @Test
public void NoUpdateTest() throws Exception { public void NoUpdateTest() throws Exception {
SparkResultToProjectThroughSemRelJob3.main(new String[]{ SparkResultToProjectThroughSemRelJob3.main(
"-isTest", Boolean.TRUE.toString(), new String[] {
"-isSparkSessionManaged", Boolean.FALSE.toString(), "-isTest",
// "-sourcePath", getClass().getResource("/eu/dnetlib/dhp/resulttoorganizationfrominstrepo/sample/relation").getPath(), Boolean.TRUE.toString(),
"-hive_metastore_uris", "", "-isSparkSessionManaged",
"-writeUpdate", "false", Boolean.FALSE.toString(),
"-saveGraph", "true", // "-sourcePath",
"-outputPath", workingDir.toString() + "/relation", // getClass().getResource("/eu/dnetlib/dhp/resulttoorganizationfrominstrepo/sample/relation").getPath(),
"-potentialUpdatePath", getClass().getResource("/eu/dnetlib/dhp/projecttoresult/preparedInfo/noupdates/potentialUpdates").getPath(), "-hive_metastore_uris",
"-alreadyLinkedPath",getClass().getResource("/eu/dnetlib/dhp/projecttoresult/preparedInfo/alreadyLinked").getPath(), "",
"-saveGraph",
"true",
"-outputPath",
workingDir.toString() + "/relation",
"-potentialUpdatePath",
getClass()
.getResource(
"/eu/dnetlib/dhp/projecttoresult/preparedInfo/noupdates/potentialUpdates")
.getPath(),
"-alreadyLinkedPath",
getClass()
.getResource(
"/eu/dnetlib/dhp/projecttoresult/preparedInfo/alreadyLinked")
.getPath(),
}); });
final JavaSparkContext sc = new JavaSparkContext(spark.sparkContext()); final JavaSparkContext sc = new JavaSparkContext(spark.sparkContext());
JavaRDD<Relation> tmp = sc.textFile(workingDir.toString()+"/relation") JavaRDD<Relation> tmp =
sc.textFile(workingDir.toString() + "/relation")
.map(item -> OBJECT_MAPPER.readValue(item, Relation.class)); .map(item -> OBJECT_MAPPER.readValue(item, Relation.class));
Assertions.assertEquals(0, tmp.count()); Assertions.assertEquals(0, tmp.count());
} }
/** /**
* All the possible updates will produce a new relation. No relations are already linked in the grpha * All the possible updates will produce a new relation. No relations are already linked in the
* grpha
*
* @throws Exception * @throws Exception
*/ */
@Test @Test
public void UpdateTenTest() throws Exception{ public void UpdateTenTest() throws Exception {
SparkResultToProjectThroughSemRelJob3.main(new String[]{ SparkResultToProjectThroughSemRelJob3.main(
"-isTest", Boolean.TRUE.toString(), new String[] {
"-isSparkSessionManaged", Boolean.FALSE.toString(), "-isTest",
// "-sourcePath", getClass().getResource("/eu/dnetlib/dhp/resulttoorganizationfrominstrepo/sample/relation").getPath(), Boolean.TRUE.toString(),
"-hive_metastore_uris", "", "-isSparkSessionManaged",
"-writeUpdate", "false", Boolean.FALSE.toString(),
"-saveGraph", "true", // "-sourcePath",
"-outputPath", workingDir.toString() + "/relation", // getClass().getResource("/eu/dnetlib/dhp/resulttoorganizationfrominstrepo/sample/relation").getPath(),
"-potentialUpdatePath", getClass().getResource("/eu/dnetlib/dhp/projecttoresult/preparedInfo/tenupdates/potentialUpdates").getPath(), "-hive_metastore_uris",
"-alreadyLinkedPath",getClass().getResource("/eu/dnetlib/dhp/projecttoresult/preparedInfo/alreadyLinked").getPath(), "",
"-saveGraph",
"true",
"-outputPath",
workingDir.toString() + "/relation",
"-potentialUpdatePath",
getClass()
.getResource(
"/eu/dnetlib/dhp/projecttoresult/preparedInfo/tenupdates/potentialUpdates")
.getPath(),
"-alreadyLinkedPath",
getClass()
.getResource(
"/eu/dnetlib/dhp/projecttoresult/preparedInfo/alreadyLinked")
.getPath(),
}); });
final JavaSparkContext sc = new JavaSparkContext(spark.sparkContext()); final JavaSparkContext sc = new JavaSparkContext(spark.sparkContext());
JavaRDD<Relation> tmp = sc.textFile(workingDir.toString()+"/relation") JavaRDD<Relation> tmp =
sc.textFile(workingDir.toString() + "/relation")
.map(item -> OBJECT_MAPPER.readValue(item, Relation.class)); .map(item -> OBJECT_MAPPER.readValue(item, Relation.class));
//got 20 new relations because "produces" and "isProducedBy" are added // got 20 new relations because "produces" and "isProducedBy" are added
Assertions.assertEquals(20, tmp.count()); Assertions.assertEquals(10, tmp.count());
Dataset<Relation> verificationDs = spark.createDataset(tmp.rdd(), Encoders.bean(Relation.class)); Dataset<Relation> verificationDs =
spark.createDataset(tmp.rdd(), Encoders.bean(Relation.class));
Assertions.assertEquals(10, verificationDs.filter("relClass = 'produces'").count()); Assertions.assertEquals(5, verificationDs.filter("relClass = 'produces'").count());
Assertions.assertEquals(10, verificationDs.filter("relClass = 'isProducedBy'").count()); Assertions.assertEquals(5, verificationDs.filter("relClass = 'isProducedBy'").count());
Assertions.assertEquals(10, verificationDs.filter(r -> r.getSource().substring(0, 2). equals("50") && Assertions.assertEquals(
r.getTarget().substring(0,2).equals("40") && 5,
r.getRelClass().equals("isProducedBy")).count()); verificationDs
Assertions.assertEquals(10, verificationDs.filter(r -> r.getSource().substring(0, 2). equals("40") && .filter(
r.getTarget().substring(0,2).equals("50") && r ->
r.getRelClass().equals("produces")).count()); r.getSource().substring(0, 2).equals("50")
&& r.getTarget().substring(0, 2).equals("40")
&& r.getRelClass().equals("isProducedBy"))
.count());
Assertions.assertEquals(
5,
verificationDs
.filter(
r ->
r.getSource().substring(0, 2).equals("40")
&& r.getTarget().substring(0, 2).equals("50")
&& r.getRelClass().equals("produces"))
.count());
verificationDs.createOrReplaceTempView("temporary"); verificationDs.createOrReplaceTempView("temporary");
Assertions.assertEquals(20, spark.sql("Select * from temporary where datainfo.inferenceprovenance = 'propagation'").count()); Assertions.assertEquals(
10,
spark.sql(
"Select * from temporary where datainfo.inferenceprovenance = 'propagation'")
.count());
} }
/** /**
* One of the relations in the possible updates is already linked to the project in the graph. * One of the relations in the possible updates is already linked to the project in the graph.
* All the others are not. There will be 9 new associations leading to 18 new relations * All the others are not. There will be 9 new associations leading to 18 new relations
*
* @throws Exception * @throws Exception
*/ */
@Test @Test
public void UpdateMixTest() throws Exception{ public void UpdateMixTest() throws Exception {
SparkResultToProjectThroughSemRelJob3.main(new String[]{ SparkResultToProjectThroughSemRelJob3.main(
"-isTest", Boolean.TRUE.toString(), new String[] {
"-isSparkSessionManaged", Boolean.FALSE.toString(), "-isTest",
// "-sourcePath", getClass().getResource("/eu/dnetlib/dhp/resulttoorganizationfrominstrepo/sample/relation").getPath(), Boolean.TRUE.toString(),
"-hive_metastore_uris", "", "-isSparkSessionManaged",
"-writeUpdate", "false", Boolean.FALSE.toString(),
"-saveGraph", "true", // "-sourcePath",
"-outputPath", workingDir.toString() + "/relation", // getClass().getResource("/eu/dnetlib/dhp/resulttoorganizationfrominstrepo/sample/relation").getPath(),
"-potentialUpdatePath", getClass().getResource("/eu/dnetlib/dhp/projecttoresult/preparedInfo/updatesmixed/potentialUpdates").getPath(), "-hive_metastore_uris",
"-alreadyLinkedPath",getClass().getResource("/eu/dnetlib/dhp/projecttoresult/preparedInfo/alreadyLinked").getPath(), "",
"-saveGraph",
"true",
"-outputPath",
workingDir.toString() + "/relation",
"-potentialUpdatePath",
getClass()
.getResource(
"/eu/dnetlib/dhp/projecttoresult/preparedInfo/updatesmixed/potentialUpdates")
.getPath(),
"-alreadyLinkedPath",
getClass()
.getResource(
"/eu/dnetlib/dhp/projecttoresult/preparedInfo/alreadyLinked")
.getPath(),
}); });
final JavaSparkContext sc = new JavaSparkContext(spark.sparkContext()); final JavaSparkContext sc = new JavaSparkContext(spark.sparkContext());
JavaRDD<Relation> tmp = sc.textFile(workingDir.toString()+"/relation") JavaRDD<Relation> tmp =
sc.textFile(workingDir.toString() + "/relation")
.map(item -> OBJECT_MAPPER.readValue(item, Relation.class)); .map(item -> OBJECT_MAPPER.readValue(item, Relation.class));
// JavaRDD<Relation> tmp = sc.textFile("/tmp/relation") // JavaRDD<Relation> tmp = sc.textFile("/tmp/relation")
// .map(item -> OBJECT_MAPPER.readValue(item, Relation.class)); // .map(item -> OBJECT_MAPPER.readValue(item, Relation.class));
// got 20 new relations because "produces" and "isProducedBy" are added
Assertions.assertEquals(8, tmp.count());
//got 20 new relations because "produces" and "isProducedBy" are added Dataset<Relation> verificationDs =
Assertions.assertEquals(18, tmp.count()); spark.createDataset(tmp.rdd(), Encoders.bean(Relation.class));
Dataset<Relation> verificationDs = spark.createDataset(tmp.rdd(), Encoders.bean(Relation.class)); Assertions.assertEquals(4, verificationDs.filter("relClass = 'produces'").count());
Assertions.assertEquals(4, verificationDs.filter("relClass = 'isProducedBy'").count());
Assertions.assertEquals(9, verificationDs.filter("relClass = 'produces'").count()); Assertions.assertEquals(
Assertions.assertEquals(9, verificationDs.filter("relClass = 'isProducedBy'").count()); 4,
verificationDs
Assertions.assertEquals(9, verificationDs.filter(r -> r.getSource().substring(0, 2). equals("50") && .filter(
r.getTarget().substring(0,2).equals("40") && r ->
r.getRelClass().equals("isProducedBy")).count()); r.getSource().substring(0, 2).equals("50")
Assertions.assertEquals(9, verificationDs.filter(r -> r.getSource().substring(0, 2). equals("40") && && r.getTarget().substring(0, 2).equals("40")
r.getTarget().substring(0,2).equals("50") && && r.getRelClass().equals("isProducedBy"))
r.getRelClass().equals("produces")).count()); .count());
Assertions.assertEquals(
4,
verificationDs
.filter(
r ->
r.getSource().substring(0, 2).equals("40")
&& r.getTarget().substring(0, 2).equals("50")
&& r.getRelClass().equals("produces"))
.count());
verificationDs.createOrReplaceTempView("temporary"); verificationDs.createOrReplaceTempView("temporary");
Assertions.assertEquals(18, spark.sql("Select * from temporary where datainfo.inferenceprovenance = 'propagation'").count()); Assertions.assertEquals(
8,
spark.sql(
"Select * from temporary where datainfo.inferenceprovenance = 'propagation'")
.count());
} }
} }