antonis.lempesis
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
1
Fork 1
Code Issues Pull Requests Projects Releases Wiki Activity

-

This commit is contained in:
Miriam Baglioni 2021-12-14 11:12:17 +01:00
parent 936578aaf1
commit 4eb8276493
8 changed files with 185 additions and 252 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

49
dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/actionmanager/bipfinder/Constants.java Normal file
View File

@ -0,0 +1,49 @@
package eu.dnetlib.dhp.actionmanager.bipfinder;
import java.util.Optional;
import org.apache.spark.api.java.function.MapFunction;
import org.apache.spark.sql.Dataset;
import org.apache.spark.sql.Encoders;
import org.apache.spark.sql.SparkSession;
import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
public class Constants {
public static final String DOI = "doi";
public static final String UPDATE_DATA_INFO_TYPE = "update";
public static final String UPDATE_SUBJECT_FOS_CLASS_ID = "subject:fos";
public static final String UPDATE_CLASS_NAME = "Inferred by OpenAIRE";
public static final String UPDATE_MEASURE_BIP_CLASS_ID = "measure:bip";
public static final String FOS_CLASS_ID = "FOS";
public static final String FOS_CLASS_NAME = "Fields of Science and Technology classification";
public static final String NULL = "NULL";
public static final ObjectMapper OBJECT_MAPPER = new ObjectMapper();
private Constants() {
}
public static Boolean isSparkSessionManaged(ArgumentApplicationParser parser) {
return Optional
.ofNullable(parser.get("isSparkSessionManaged"))
.map(Boolean::valueOf)
.orElse(Boolean.TRUE);
}
public static <R> Dataset<R> readPath(
SparkSession spark, String inputPath, Class<R> clazz) {
return spark
.read()
.textFile(inputPath)
.map((MapFunction<String, R>) value -> OBJECT_MAPPER.readValue(value, clazz), Encoders.bean(clazz));
}
}

28
dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/actionmanager/bipfinder/PreparedResult.java
View File

@ -1,28 +0,0 @@
package eu.dnetlib.dhp.actionmanager.bipfinder;
import java.io.Serializable;
/**
* Subset of the information of the generic results that are needed to create the atomic action
*/
public class PreparedResult implements Serializable {
private String id; // openaire id
private String value; // doi
public String getId() {
return id;
}
public void setId(String id) {
this.id = id;
}
public String getValue() {
return value;
}
public void setValue(String value) {
this.value = value;
}
}

64
dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/actionmanager/bipfinder/SparkAtomicActionScoreJob.java
View File

@ -1,6 +1,7 @@
package eu.dnetlib.dhp.actionmanager.bipfinder;
import static eu.dnetlib.dhp.actionmanager.createunresolvedentities.Constants.*;
import static eu.dnetlib.dhp.common.SparkSessionSupport.runWithSparkSession;
import java.io.Serializable;
@ -18,6 +19,7 @@ import org.apache.spark.api.java.function.MapFunction;
import org.apache.spark.api.java.function.MapGroupsFunction;
import org.apache.spark.sql.Dataset;
import org.apache.spark.sql.Encoders;
import org.apache.spark.sql.Row;
import org.apache.spark.sql.SparkSession;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
@ -27,8 +29,10 @@ import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
import eu.dnetlib.dhp.common.HdfsSupport;
import eu.dnetlib.dhp.schema.action.AtomicAction;
import eu.dnetlib.dhp.schema.common.ModelConstants;
import eu.dnetlib.dhp.schema.oaf.*;
import eu.dnetlib.dhp.schema.oaf.KeyValue;
import eu.dnetlib.dhp.schema.oaf.utils.OafMapperUtils;
import scala.Tuple2;
/**
@ -105,32 +109,29 @@ public class SparkAtomicActionScoreJob implements Serializable {
results.createOrReplaceTempView("result");
Dataset<PreparedResult> preparedResult = spark
Dataset<Row> preparedResult = spark
.sql(
"select pIde.value value, id " +
"select id " +
"from result " +
"lateral view explode (pid) p as pIde " +
"where dataInfo.deletedbyinference = false and pIde.qualifier.classid = '" + DOI + "'")
.as(Encoders.bean(PreparedResult.class));
"where dataInfo.deletedbyinference = false ");
bipScores
.joinWith(
preparedResult, bipScores.col("id").equalTo(preparedResult.col("value")),
preparedResult, bipScores.col("id").equalTo(preparedResult.col("id")),
"inner")
.map((MapFunction<Tuple2<BipScore, PreparedResult>, BipScore>) value -> {
.map((MapFunction<Tuple2<BipScore, Row>, BipScore>) value -> {
BipScore ret = value._1();
ret.setId(value._2().getId());
ret.setId(value._1().getId());
return ret;
}, Encoders.bean(BipScore.class))
.groupByKey((MapFunction<BipScore, String>) BipScore::getId, Encoders.STRING())
.mapGroups((MapGroupsFunction<String, BipScore, Result>) (k, it) -> {
Result ret = new Result();
ret.setDataInfo(getDataInfo());
BipScore first = it.next();
ret.setId(first.getId());
ret.setMeasures(getMeasure(first));
it.forEachRemaining(value -> ret.getMeasures().addAll(getMeasure(value)));
.map((MapFunction<BipScore, Result>) bs -> {
Result ret = new Result();
ret.setId(bs.getId());
ret.setMeasures(getMeasure(bs));
return ret;
}, Encoders.bean(Result.class))
@ -159,7 +160,21 @@ public class SparkAtomicActionScoreJob implements Serializable {
KeyValue kv = new KeyValue();
kv.setValue(unit.getValue());
kv.setKey(unit.getKey());
kv.setDataInfo(getDataInfo());
kv
.setDataInfo(
OafMapperUtils
.dataInfo(
false,
UPDATE_DATA_INFO_TYPE,
true,
false,
OafMapperUtils
.qualifier(
UPDATE_MEASURE_BIP_CLASS_ID,
UPDATE_CLASS_NAME,
ModelConstants.DNET_PROVENANCE_ACTIONS,
ModelConstants.DNET_PROVENANCE_ACTIONS),
""));
return kv;
})
.collect(Collectors.toList()));
@ -168,21 +183,6 @@ public class SparkAtomicActionScoreJob implements Serializable {
.collect(Collectors.toList());
}
private static DataInfo getDataInfo() {
DataInfo di = new DataInfo();
di.setInferred(false);
di.setInvisible(false);
di.setDeletedbyinference(false);
di.setTrust("");
Qualifier qualifier = new Qualifier();
qualifier.setClassid("sysimport:actionset");
qualifier.setClassname("Harvested");
qualifier.setSchemename("dnet:provenanceActions");
qualifier.setSchemeid("dnet:provenanceActions");
di.setProvenanceaction(qualifier);
return di;
}
private static void removeOutputDir(SparkSession spark, String path) {
HdfsSupport.remove(path, spark.sparkContext().hadoopConfiguration());
}

30
dhp-workflows/dhp-aggregation/src/main/resources/eu/dnetlib/dhp/actionmanager/bipfinder/oozie_app/config-default.xml Normal file
View File

@ -0,0 +1,30 @@
<configuration>
<property>
<name>jobTracker</name>
<value>yarnRM</value>
</property>
<property>
<name>nameNode</name>
<value>hdfs://nameservice1</value>
</property>
<property>
<name>oozie.use.system.libpath</name>
<value>true</value>
</property>
<property>
<name>hiveMetastoreUris</name>
<value>thrift://iis-cdh5-test-m3.ocean.icm.edu.pl:9083</value>
</property>
<property>
<name>hiveJdbcUrl</name>
<value>jdbc:hive2://iis-cdh5-test-m3.ocean.icm.edu.pl:10000</value>
</property>
<property>
<name>hiveDbName</name>
<value>openaire</value>
</property>
<property>
<name>oozie.launcher.mapreduce.user.classpath.first</name>
<value>true</value>
</property>
</configuration>

34
dhp-workflows/dhp-aggregation/src/main/resources/eu/dnetlib/dhp/actionmanager/bipfinder/oozie_app/workflow.xml
View File

@ -13,6 +13,40 @@
<name>outputPath</name>
<description>the path where to store the actionset</description>
</property>
<property>
<name>sparkDriverMemory</name>
<description>memory for driver process</description>
</property>
<property>
<name>sparkExecutorMemory</name>
<description>memory for individual executor</description>
</property>
<property>
<name>sparkExecutorCores</name>
<description>number of cores used by single executor</description>
</property>
<property>
<name>oozieActionShareLibForSpark2</name>
<description>oozie action sharelib for spark 2.*</description>
</property>
<property>
<name>spark2ExtraListeners</name>
<value>com.cloudera.spark.lineage.NavigatorAppListener</value>
<description>spark 2.* extra listeners classname</description>
</property>
<property>
<name>spark2SqlQueryExecutionListeners</name>
<value>com.cloudera.spark.lineage.NavigatorQueryListener</value>
<description>spark 2.* sql query execution listeners classname</description>
</property>
<property>
<name>spark2YarnHistoryServerAddress</name>
<description>spark 2.* yarn history server address</description>
</property>
<property>
<name>spark2EventLogDir</name>
<description>spark 2.* event log dir location</description>
</property>
</parameters>
<start to="deleteoutputpath"/>

210
dhp-workflows/dhp-aggregation/src/test/java/eu/dnetlib/dhp/actionmanager/bipfinder/SparkAtomicActionScoreJobTest.java
View File

@ -28,6 +28,7 @@ import com.fasterxml.jackson.databind.ObjectMapper;
import eu.dnetlib.dhp.schema.action.AtomicAction;
import eu.dnetlib.dhp.schema.oaf.Publication;
import eu.dnetlib.dhp.schema.oaf.Result;
public class SparkAtomicActionScoreJobTest {
@ -69,13 +70,13 @@ public class SparkAtomicActionScoreJobTest {
}
@Test
void matchOne() throws Exception {
void testMatch() throws Exception {
String bipScoresPath = getClass()
.getResource("/eu/dnetlib/dhp/actionmanager/bipfinder/bip_scores.json")
.getResource("/eu/dnetlib/dhp/actionmanager/bipfinder/bip_scores_oid.json")
.getPath();
String inputPath = getClass()
.getResource(
"/eu/dnetlib/dhp/actionmanager/bipfinder/publication.json")
"/eu/dnetlib/dhp/actionmanager/bipfinder/publicationoaid.json")
.getPath();
SparkAtomicActionScoreJob
@ -95,223 +96,48 @@ public class SparkAtomicActionScoreJobTest {
final JavaSparkContext sc = new JavaSparkContext(spark.sparkContext());
JavaRDD<Publication> tmp = sc
JavaRDD<Result> tmp = sc
.sequenceFile(workingDir.toString() + "/actionSet", Text.class, Text.class)
.map(value -> OBJECT_MAPPER.readValue(value._2().toString(), AtomicAction.class))
.map(aa -> ((Publication) aa.getPayload()));
.map(aa -> ((Result) aa.getPayload()));
assertEquals(1, tmp.count());
assertEquals(4, tmp.count());
Dataset<Publication> verificationDataset = spark.createDataset(tmp.rdd(), Encoders.bean(Publication.class));
verificationDataset.createOrReplaceTempView("publication");
Dataset<Result> verificationDataset = spark.createDataset(tmp.rdd(), Encoders.bean(Result.class));
verificationDataset.createOrReplaceTempView("result");
Dataset<Row> execVerification = spark
.sql(
"Select p.id oaid, mes.id, mUnit.value from publication p " +
"Select p.id oaid, mes.id, mUnit.value from result p " +
"lateral view explode(measures) m as mes " +
"lateral view explode(mes.unit) u as mUnit ");
Assertions.assertEquals(2, execVerification.count());
Assertions.assertEquals(12, execVerification.count());
Assertions
.assertEquals(
"50|355e65625b88::ffa5bad14f4adc0c9a15c00efbbccddb",
execVerification.select("oaid").collectAsList().get(0).getString(0));
Assertions
.assertEquals(
"1.47565045883e-08",
execVerification.filter("id = 'influence'").select("value").collectAsList().get(0).getString(0));
Assertions
.assertEquals(
"0.227515392",
execVerification.filter("id = 'popularity'").select("value").collectAsList().get(0).getString(0));
}
@Test
void matchOneWithTwo() throws Exception {
String bipScoresPath = getClass()
.getResource("/eu/dnetlib/dhp/actionmanager/bipfinder/bip_scores.json")
.getPath();
String inputPath = getClass()
.getResource(
"/eu/dnetlib/dhp/actionmanager/bipfinder/publication_2.json")
.getPath();
SparkAtomicActionScoreJob
.main(
new String[] {
"-isSparkSessionManaged",
Boolean.FALSE.toString(),
"-inputPath",
inputPath,
"-bipScorePath",
bipScoresPath,
"-resultTableName",
"eu.dnetlib.dhp.schema.oaf.Publication",
"-outputPath",
workingDir.toString() + "/actionSet"
});
final JavaSparkContext sc = new JavaSparkContext(spark.sparkContext());
JavaRDD<Publication> tmp = sc
.sequenceFile(workingDir.toString() + "/actionSet", Text.class, Text.class)
.map(value -> OBJECT_MAPPER.readValue(value._2().toString(), AtomicAction.class))
.map(aa -> ((Publication) aa.getPayload()));
assertEquals(1, tmp.count());
Dataset<Publication> verificationDataset = spark.createDataset(tmp.rdd(), Encoders.bean(Publication.class));
verificationDataset.createOrReplaceTempView("publication");
Dataset<Row> execVerification = spark
.sql(
"Select p.id oaid, mes.id, mUnit.value from publication p " +
"lateral view explode(measures) m as mes " +
"lateral view explode(mes.unit) u as mUnit ");
Assertions.assertEquals(4, execVerification.count());
Assertions
.assertEquals(
"50|355e65625b88::ffa5bad14f4adc0c9a15c00efbbccddb",
execVerification.select("oaid").collectAsList().get(0).getString(0));
Assertions
.assertEquals(
2,
execVerification.filter("id = 'influence'").count());
Assertions
.assertEquals(
2,
execVerification.filter("id = 'popularity'").count());
List<Row> tmp_ds = execVerification.filter("id = 'influence'").select("value").collectAsList();
String tmp_influence = tmp_ds.get(0).getString(0);
assertTrue(
"1.47565045883e-08".equals(tmp_influence) ||
"1.98956540239e-08".equals(tmp_influence));
tmp_influence = tmp_ds.get(1).getString(0);
assertTrue(
"1.47565045883e-08".equals(tmp_influence) ||
"1.98956540239e-08".equals(tmp_influence));
assertNotEquals(tmp_ds.get(1).getString(0), tmp_ds.get(0).getString(0));
}
@Test
void matchTwo() throws Exception {
String bipScoresPath = getClass()
.getResource("/eu/dnetlib/dhp/actionmanager/bipfinder/bip_scores.json")
.getPath();
String inputPath = getClass()
.getResource(
"/eu/dnetlib/dhp/actionmanager/bipfinder/publication_3.json")
.getPath();
SparkAtomicActionScoreJob
.main(
new String[] {
"-isSparkSessionManaged",
Boolean.FALSE.toString(),
"-inputPath",
inputPath,
"-bipScorePath",
bipScoresPath,
"-resultTableName",
"eu.dnetlib.dhp.schema.oaf.Publication",
"-outputPath",
workingDir.toString() + "/actionSet"
});
final JavaSparkContext sc = new JavaSparkContext(spark.sparkContext());
JavaRDD<Publication> tmp = sc
.sequenceFile(workingDir.toString() + "/actionSet", Text.class, Text.class)
.map(value -> OBJECT_MAPPER.readValue(value._2().toString(), AtomicAction.class))
.map(aa -> ((Publication) aa.getPayload()));
assertEquals(2, tmp.count());
Dataset<Publication> verificationDataset = spark.createDataset(tmp.rdd(), Encoders.bean(Publication.class));
verificationDataset.createOrReplaceTempView("publication");
Dataset<Row> execVerification = spark
.sql(
"Select p.id oaid, mes.id, mUnit.value from publication p " +
"lateral view explode(measures) m as mes " +
"lateral view explode(mes.unit) u as mUnit ");
Assertions.assertEquals(4, execVerification.count());
Assertions
.assertEquals(
2,
execVerification.filter("oaid = '50|355e65625b88::ffa5bad14f4adc0c9a15c00efbbccddb'").count());
Assertions
.assertEquals(
2,
execVerification.filter("oaid = '50|acm_________::faed5b7a1bd8f51118d13ed29cfaee09'").count());
Assertions
.assertEquals(
2,
execVerification.filter("id = 'influence'").count());
Assertions
.assertEquals(
2,
execVerification.filter("id = 'popularity'").count());
Assertions
.assertEquals(
"1.47565045883e-08",
execVerification
"6.63451994567e-09", execVerification
.filter(
"oaid = '50|355e65625b88::ffa5bad14f4adc0c9a15c00efbbccddb' " +
"oaid='50|arXiv_dedup_::4a2d5fd8d71daec016c176ec71d957b1' " +
"and id = 'influence'")
.select("value")
.collectAsList()
.get(0)
.getString(0));
Assertions
.assertEquals(
"1.98956540239e-08",
execVerification
"0.348694533145", execVerification
.filter(
"oaid = '50|acm_________::faed5b7a1bd8f51118d13ed29cfaee09' " +
"and id = 'influence'")
"oaid='50|arXiv_dedup_::4a2d5fd8d71daec016c176ec71d957b1' " +
"and id = 'popularity_alt'")
.select("value")
.collectAsList()
.get(0)
.getString(0));
Assertions
.assertEquals(
"0.282046161584",
execVerification
"2.16094680115e-09", execVerification
.filter(
"oaid = '50|acm_________::faed5b7a1bd8f51118d13ed29cfaee09' " +
"and id = 'popularity'")
.select("value")
.collectAsList()
.get(0)
.getString(0));
Assertions
.assertEquals(
"0.227515392",
execVerification
.filter(
"oaid = '50|355e65625b88::ffa5bad14f4adc0c9a15c00efbbccddb' " +
"oaid='50|arXiv_dedup_::4a2d5fd8d71daec016c176ec71d957b1' " +
"and id = 'popularity'")
.select("value")
.collectAsList()

4
dhp-workflows/dhp-aggregation/src/test/resources/eu/dnetlib/dhp/actionmanager/bipfinder/bip_scores_oid.json Normal file
View File

@ -0,0 +1,4 @@
{"50|arXiv_dedup_::4a2d5fd8d71daec016c176ec71d957b1": [{"id": "influence", "unit": [{"value": "6.63451994567e-09", "key": "score"}]}, {"id": "popularity_alt", "unit": [{"value": "0.348694533145", "key": "score"}]}, {"id": "popularity", "unit": [{"value": "2.16094680115e-09", "key": "score"}]}]}
{"50|dedup_wf_001::05b1f8ce98702f69d07aa5f0429de1e3": [{"id": "influence", "unit": [{"value": "6.25057357279e-09", "key": "score"}]}, {"id": "popularity_alt", "unit": [{"value": "7.0208", "key": "score"}]}, {"id": "popularity", "unit": [{"value": "2.40234462244e-08", "key": "score"}]}]}
{"50|dedup_wf_001::08823c8f5c3ca2eae523817036cdda67": [{"id": "influence", "unit": [{"value": "5.54921449123e-09", "key": "score"}]}, {"id": "popularity_alt", "unit": [{"value": "0.0", "key": "score"}]}, {"id": "popularity", "unit": [{"value": "3.53012887452e-10", "key": "score"}]}]}
{"50|dedup_wf_001::0e72b399325d6efcbe3271891a1dfe4c": [{"id": "influence", "unit": [{"value": "1.63466096315e-08", "key": "score"}]}, {"id": "popularity_alt", "unit": [{"value": "20.9870879741", "key": "score"}]}, {"id": "popularity", "unit": [{"value": "5.49501495323e-08", "key": "score"}]}]}

18
dhp-workflows/dhp-aggregation/src/test/resources/eu/dnetlib/dhp/actionmanager/bipfinder/publicationoaid.json Normal file
View File

File diff suppressed because one or more lines are too long