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

Merge branch 'beta' into apc_affiliation

This commit is contained in:
Claudio Atzori 2023-05-15 15:47:33 +02:00
parent 8a463cc3e8 0c314d5e09
commit 8acad52a0c
20 changed files with 664 additions and 596 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/actionmanager/opencitations/CreateActionSetSparkJob.java
View File

@ -114,10 +114,10 @@ public class CreateActionSetSparkJob implements Serializable {
if (!citing.equals(cited)) {
relationList
.addAll(
getRelations(
.add(
getRelation(
citing,
cited));
cited, ModelConstants.CITES));
if (duplicate && value.getCiting().endsWith(".refs")) {
citing = ID_PREFIX + IdentifierFactory
@ -125,7 +125,7 @@ public class CreateActionSetSparkJob implements Serializable {
CleaningFunctions
.normalizePidValue(
"doi", value.getCiting().substring(0, value.getCiting().indexOf(".refs"))));
relationList.addAll(getRelations(citing, cited));
relationList.add(getRelation(citing, cited, ModelConstants.CITES));
}
}

22
dhp-workflows/dhp-aggregation/src/test/java/eu/dnetlib/dhp/actionmanager/opencitations/CreateOpenCitationsASTest.java
View File

@ -99,7 +99,7 @@ public class CreateOpenCitationsASTest {
.map(value -> OBJECT_MAPPER.readValue(value._2().toString(), AtomicAction.class))
.map(aa -> ((Relation) aa.getPayload()));
assertEquals(62, tmp.count());
assertEquals(31, tmp.count());
// tmp.foreach(r -> System.out.println(OBJECT_MAPPER.writeValueAsString(r)));
@ -131,7 +131,7 @@ public class CreateOpenCitationsASTest {
.map(value -> OBJECT_MAPPER.readValue(value._2().toString(), AtomicAction.class))
.map(aa -> ((Relation) aa.getPayload()));
assertEquals(46, tmp.count());
assertEquals(23, tmp.count());
// tmp.foreach(r -> System.out.println(OBJECT_MAPPER.writeValueAsString(r)));
@ -241,7 +241,7 @@ public class CreateOpenCitationsASTest {
assertEquals("resultResult", r.getRelType());
});
assertEquals(23, tmp.filter(r -> r.getRelClass().equals("Cites")).count());
assertEquals(23, tmp.filter(r -> r.getRelClass().equals("IsCitedBy")).count());
assertEquals(0, tmp.filter(r -> r.getRelClass().equals("IsCitedBy")).count());
}
@ -318,15 +318,15 @@ public class CreateOpenCitationsASTest {
JavaRDD<Relation> check = tmp.filter(r -> r.getSource().equals(doi1) || r.getTarget().equals(doi1));
assertEquals(10, check.count());
assertEquals(5, check.count());
check.foreach(r -> {
if (r.getSource().equals(doi2) || r.getSource().equals(doi3) || r.getSource().equals(doi4) ||
r.getSource().equals(doi5) || r.getSource().equals(doi6)) {
assertEquals(ModelConstants.IS_CITED_BY, r.getRelClass());
assertEquals(doi1, r.getTarget());
}
});
// check.foreach(r -> {
// if (r.getSource().equals(doi2) || r.getSource().equals(doi3) || r.getSource().equals(doi4) ||
// r.getSource().equals(doi5) || r.getSource().equals(doi6)) {
// assertEquals(ModelConstants.IS_CITED_BY, r.getRelClass());
// assertEquals(doi1, r.getTarget());
// }
// });
assertEquals(5, check.filter(r -> r.getSource().equals(doi1)).count());
check.filter(r -> r.getSource().equals(doi1)).foreach(r -> assertEquals(ModelConstants.CITES, r.getRelClass()));

45
dhp-workflows/dhp-enrichment/src/main/java/eu/dnetlib/dhp/resulttocommunityfromorganization/PrepareResultCommunitySet.java
View File

@ -72,33 +72,28 @@ public class PrepareResultCommunitySet {
String outputPath,
OrganizationMap organizationMap) {
Dataset<Relation> relationAffiliation = readPath(spark, inputPath, Relation.class)
.filter(
(FilterFunction<Relation>) r -> !r.getDataInfo().getDeletedbyinference() &&
r.getRelClass().equalsIgnoreCase(ModelConstants.HAS_AUTHOR_INSTITUTION));
Dataset<Relation> relation = readPath(spark, inputPath, Relation.class);
relation.createOrReplaceTempView("relation");
Dataset<Relation> relationOrganization = readPath(spark, inputPath, Relation.class)
.filter(
(FilterFunction<Relation>) r -> !r.getDataInfo().getDeletedbyinference() &&
r.getRelClass().equalsIgnoreCase(ModelConstants.MERGES));
String query = "SELECT result_organization.source resultId, result_organization.target orgId, org_set merges "
+ "FROM (SELECT source, target "
+ " FROM relation "
+ " WHERE datainfo.deletedbyinference = false "
+ " AND lower(relClass) = '"
+ ModelConstants.HAS_AUTHOR_INSTITUTION.toLowerCase()
+ "') result_organization "
+ "LEFT JOIN (SELECT source, collect_set(target) org_set "
+ " FROM relation "
+ " WHERE datainfo.deletedbyinference = false "
+ " AND lower(relClass) = '"
+ ModelConstants.MERGES.toLowerCase()
+ "' "
+ " GROUP BY source) organization_organization "
+ "ON result_organization.target = organization_organization.source ";
Dataset<ResultOrganizations> result_organizationset = relationAffiliation
.joinWith(
relationOrganization,
relationAffiliation.col("target").equalTo(relationOrganization.col("source")),
"left")
.groupByKey((MapFunction<Tuple2<Relation, Relation>, String>) t2 -> t2._2().getSource(), Encoders.STRING())
.mapGroups((MapGroupsFunction<String, Tuple2<Relation, Relation>, ResultOrganizations>) (k, it) -> {
ResultOrganizations rOrgs = new ResultOrganizations();
rOrgs.setOrgId(k);
Tuple2<Relation, Relation> first = it.next();
rOrgs.setResultId(first._1().getSource());
ArrayList<String> merges = new ArrayList<>();
merges.add(first._2().getTarget());
it.forEachRemaining(t -> merges.add(t._2().getTarget()));
rOrgs.setMerges(merges);
return rOrgs;
}, Encoders.bean(ResultOrganizations.class));
Dataset<ResultOrganizations> result_organizationset = spark
.sql(query)
.as(Encoders.bean(ResultOrganizations.class));
result_organizationset
.map(mapResultCommunityFn(organizationMap), Encoders.bean(ResultCommunityList.class))

2
dhp-workflows/dhp-graph-mapper/src/main/java/eu/dnetlib/dhp/oa/graph/clean/GetDatasourceFromCountry.java
View File

@ -78,6 +78,8 @@ public class GetDatasourceFromCountry implements Serializable {
Encoders.bean(Organization.class))
.filter(
(FilterFunction<Organization>) o -> !o.getDataInfo().getDeletedbyinference() &&
o.getCountry() != null &&
o.getCountry().getClassid() != null &&
o.getCountry().getClassid().length() > 0 &&
o.getCountry().getClassid().equals(country));

23
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/contexts.sh
View File

@ -9,6 +9,8 @@ fi
CONTEXT_API=$1
TARGET_DB=$2
export HIVE_OPTS="-hiveconf mapred.job.queue.name=analytics -hiveconf hive.spark.client.connect.timeout=120000ms -hiveconf hive.spark.client.server.connect.timeout=300000ms -hiveconf spark.executor.memory=4831838208 -hiveconf spark.yarn.executor.memoryOverhead=450"
TMP=/tmp/stats-update-`tr -dc A-Za-z0-9 </dev/urandom | head -c 6`
echo "Downloading context ids"
@ -29,21 +31,20 @@ hdfs dfs -copyFromLocal categories.csv ${TMP}
hdfs dfs -copyFromLocal concepts.csv ${TMP}
hdfs dfs -chmod -R 777 ${TMP}
export HADOOP_USER="oozie"
export HADOOP_USER_NAME="oozie"
echo "Creating and populating impala tables"
impala-shell -q "invalidate metadata"
impala-shell -d ${TARGET_DB} -q "invalidate metadata"
impala-shell -q "create table ${TARGET_DB}.context (id string, name string) row format delimited fields terminated by ','"
impala-shell -q "create table ${TARGET_DB}.category (context string, id string, name string) row format delimited fields terminated by ','"
impala-shell -q "create table ${TARGET_DB}.concept (category string, id string, name string) row format delimited fields terminated by ','"
impala-shell -d ${TARGET_DB} -q "invalidate metadata"
impala-shell -q "load data inpath '${TMP}/contexts.csv' into table ${TARGET_DB}.context"
impala-shell -q "load data inpath '${TMP}/categories.csv' into table ${TARGET_DB}.category"
impala-shell -q "load data inpath '${TMP}/concepts.csv' into table ${TARGET_DB}.concept"
hive $HIVE_OPTS -e "create table ${TARGET_DB}.context (id string, name string) row format delimited fields terminated by ','"
hive $HIVE_OPTS -e "create table ${TARGET_DB}.category (context string, id string, name string) row format delimited fields terminated by ','"
hive $HIVE_OPTS -e "create table ${TARGET_DB}.concept (category string, id string, name string) row format delimited fields terminated by ','"
hive $HIVE_OPTS -e "load data inpath '${TMP}/contexts.csv' into table ${TARGET_DB}.context"
hive $HIVE_OPTS -e "load data inpath '${TMP}/categories.csv' into table ${TARGET_DB}.category"
hive $HIVE_OPTS -e "load data inpath '${TMP}/concepts.csv' into table ${TARGET_DB}.concept"
echo "Cleaning up"
hdfs dfs -rm -f -r -skipTrash ${TMP}
rm concepts.csv
rm categories.csv
rm contexts.csv
echo "Finito!"
echo "Finito!"

63
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/copyDataToImpalaCluster.sh Normal file
View File

@ -0,0 +1,63 @@
export PYTHON_EGG_CACHE=/home/$(whoami)/.python-eggs
export link_folder=/tmp/impala-shell-python-egg-cache-$(whoami)
if ! [ -L $link_folder ]
then
rm -Rf "$link_folder"
ln -sfn ${PYTHON_EGG_CACHE}${link_folder} ${link_folder}
fi
#export HADOOP_USER_NAME="dimitris.pierrakos"
export HADOOP_USER_NAME=$5
function copydb() {
db=$1
FILE=("hive_wf_tmp_"$RANDOM)
hdfs dfs -mkdir hdfs://impala-cluster-mn1.openaire.eu:8020/tmp/$FILE/
# copy the databases from ocean to impala
echo "copying $db"
hadoop distcp -Dmapreduce.map.memory.mb=6144 -pb hdfs://nameservice1/user/hive/warehouse/${db}.db hdfs://impala-cluster-mn1.openaire.eu:8020/tmp/$FILE/
# change ownership to impala
hdfs dfs -conf /etc/impala_cluster/hdfs-site.xml -chmod -R 777 /tmp/$FILE/${db}.db
# create the databases
impala-shell --user $HADOOP_USER_NAME -i impala-cluster-dn1.openaire.eu -q "drop database if exists ${db} cascade";
impala-shell --user $HADOOP_USER_NAME -i impala-cluster-dn1.openaire.eu -q "create database ${db}";
impala-shell --user $HADOOP_USER_NAME -q "INVALIDATE METADATA"
echo "creating schema for ${db}"
for i in `impala-shell --user $HADOOP_USER_NAME -d ${db} --delimited -q "show tables"`;
do
impala-shell --user $HADOOP_USER_NAME -d ${db} --delimited -q "show create table $i";
done | sed 's/"$/;/' | sed 's/^"//' | sed 's/[[:space:]]\date[[:space:]]/`date`/g' | impala-shell --user $HADOOP_USER_NAME -i impala-cluster-dn1.openaire.eu -c -f -
# run the same command twice because we may have failures in the first run (due to views pointing to the same db)
for i in `impala-shell --user $HADOOP_USER_NAME -d ${db} --delimited -q "show tables"`;
do
impala-shell --user $HADOOP_USER_NAME -d ${db} --delimited -q "show create table $i";
done | sed 's/"$/;/' | sed 's/^"//' | sed 's/[[:space:]]\date[[:space:]]/`date`/g' | impala-shell --user $HADOOP_USER_NAME -i impala-cluster-dn1.openaire.eu -c -f -
# load the data from /tmp in the respective tables
echo "copying data in tables and computing stats"
for i in `impala-shell --user $HADOOP_USER_NAME -i impala-cluster-dn1.openaire.eu -d ${db} --delimited -q "show tables"`;
do
impala-shell --user $HADOOP_USER_NAME -i impala-cluster-dn1.openaire.eu -d ${db} -q "load data inpath '/tmp/$FILE/${db}.db/$i' into table $i";
impala-shell --user $HADOOP_USER_NAME -i impala-cluster-dn1.openaire.eu -d ${db} -q "compute stats $i";
done
# deleting the remaining directory from hdfs
hdfs dfs -conf /etc/impala_cluster/hdfs-site.xml -rm -R /tmp/$FILE/${db}.db
}
STATS_DB=$1
MONITOR_DB=$2
OBSERVATORY_DB=$3
EXT_DB=$4
HADOOP_USER_NAME=$5
copydb $EXT_DB
copydb $STATS_DB
copydb $MONITOR_DB
copydb $OBSERVATORY_DB

28
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/finalizeImpalaCluster.sh Normal file
View File

@ -0,0 +1,28 @@
export PYTHON_EGG_CACHE=/home/$(whoami)/.python-eggs
export link_folder=/tmp/impala-shell-python-egg-cache-$(whoami)
if ! [ -L $link_folder ]
then
rm -Rf "$link_folder"
ln -sfn ${PYTHON_EGG_CACHE}${link_folder} ${link_folder}
fi
function createShadowDB() {
SOURCE=$1
SHADOW=$2
impala-shell -i impala-cluster-dn1.openaire.eu -q "drop database ${SHADOW} CASCADE";
impala-shell -i impala-cluster-dn1.openaire.eu -q "create database if not exists ${SHADOW}";
# impala-shell -i impala-cluster-dn1.openaire.eu -d ${SHADOW} -q "show tables" | sed "s/^/drop view if exists ${SHADOW}./" | sed "s/$/;/" | impala-shell -i impala-cluster-dn1.openaire.eu -f -
impala-shell -i impala-cluster-dn1.openaire.eu -d ${SOURCE} -q "show tables" --delimited | sed "s/\(.*\)/create view ${SHADOW}.\1 as select * from ${SOURCE}.\1;/" | impala-shell -i impala-cluster-dn1.openaire.eu -f -
}
STATS_DB=$1
STATS_DB_SHADOW=$2
MONITOR_DB=$3
MONITOR_DB_SHADOW=$4
OBSERVATORY_DB=$5
OBSERVATORY_DB_SHADOW=$6
createShadowDB $STATS_DB $STATS_DB_SHADOW
createShadowDB $MONITOR_DB $MONITOR_DB_SHADOW
createShadowDB $OBSERVATORY_DB $OBSERVATORY_DB_SHADOW

14
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/finalizedb.sh
View File

@ -8,12 +8,12 @@ fi
export SOURCE=$1
export SHADOW=$2
export HIVE_OPTS="-hiveconf mapred.job.queue.name=analytics -hiveconf hive.spark.client.connect.timeout=120000ms -hiveconf hive.spark.client.server.connect.timeout=300000ms -hiveconf spark.executor.memory=19166291558 -hiveconf spark.yarn.executor.memoryOverhead=3225 -hiveconf spark.driver.memory=11596411699 -hiveconf spark.yarn.driver.memoryOverhead=1228"
export HADOOP_USER_NAME="oozie"
echo "Updating shadow database"
impala-shell -q "invalidate metadata"
impala-shell -d ${SOURCE} -q "invalidate metadata"
impala-shell -d ${SOURCE} -q "show tables" --delimited | sed "s/^\(.*\)/compute stats ${SOURCE}.\1;/" | impala-shell -c -f -
impala-shell -q "create database if not exists ${SHADOW}"
impala-shell -d ${SHADOW} -q "show tables" --delimited | sed "s/^/drop view if exists ${SHADOW}./" | sed "s/$/;/" | impala-shell -c -f -
impala-shell -d ${SOURCE} -q "show tables" --delimited | sed "s/\(.*\)/create view ${SHADOW}.\1 as select * from ${SOURCE}.\1;/" | impala-shell -c -f -
echo "Shadow db ready!"
hive -e "drop database if exists ${SHADOW} cascade"
hive -e "create database if not exists ${SHADOW}"
hive $HIVE_OPTS --database ${SOURCE} -e "show tables" | grep -v WARN | sed "s/\(.*\)/create view ${SHADOW}.\1 as select * from ${SOURCE}.\1;/" > foo
hive -f foo
echo "Updated shadow database"

14
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/indicators.sh
View File

@ -7,13 +7,17 @@ then
fi
export TARGET=$1
export SCRIPT_PATH=$2
export STATS_EXT=$2
export SCRIPT_PATH=$3
export HIVE_OPTS="-hiveconf mapred.job.queue.name=analytics -hiveconf hive.spark.client.connect.timeout=120000ms -hiveconf hive.spark.client.server.connect.timeout=300000ms -hiveconf spark.executor.memory=19166291558 -hiveconf spark.yarn.executor.memoryOverhead=3225 -hiveconf spark.driver.memory=11596411699 -hiveconf spark.yarn.driver.memoryOverhead=1228 -hiveconf hive.auto.convert.join=false"
export HADOOP_USER_NAME="oozie"
echo "Getting file from " $SCRIPT_PATH
hdfs dfs -copyToLocal $SCRIPT_PATH
echo "Creating indicators"
impala-shell -q "invalidate metadata"
impala-shell -d ${TARGET} -q "show tables" --delimited | sed "s/^\(.*\)/compute stats ${TARGET}.\1;/" | impala-shell -c -f -
cat step16-createIndicatorsTables.sql | impala-shell -d $TARGET -f -
echo "Indicators created"
hive $HIVE_OPTS --database ${TARGET} -e "show tables" | grep -v WARN | sed "s/STATS_EXT/${STATS_EXT}/g" |sed "s/^\(.*\)/analyze table ${TARGET}.\1 compute statistics;/" > foo
hive $HIVE_OPTS -f foo
hive $HIVE_OPTS --database ${TARGET} -f step16-createIndicatorsTables.sql
echo "Indicators created"

19
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/monitor-post.sh Normal file
View File

@ -0,0 +1,19 @@
export PYTHON_EGG_CACHE=/home/$(whoami)/.python-eggs
export link_folder=/tmp/impala-shell-python-egg-cache-$(whoami)
if ! [ -L $link_folder ]
then
rm -Rf "$link_folder"
ln -sfn ${PYTHON_EGG_CACHE}${link_folder} ${link_folder}
fi
export SOURCE=$1
export SHADOW=$2
export HIVE_OPTS="-hiveconf mapred.job.queue.name=analytics -hiveconf hive.spark.client.connect.timeout=120000ms -hiveconf hive.spark.client.server.connect.timeout=300000ms -hiveconf spark.executor.memory=19166291558 -hiveconf spark.yarn.executor.memoryOverhead=3225 -hiveconf spark.driver.memory=11596411699 -hiveconf spark.yarn.driver.memoryOverhead=1228"
export HADOOP_USER_NAME="oozie"
echo "Updating shadow database"
hive -e "drop database if exists ${SHADOW} cascade"
hive -e "create database if not exists ${SHADOW}"
hive $HIVE_OPTS --database ${SOURCE} -e "show tables" | grep -v WARN | sed "s/\(.*\)/create view ${SHADOW}.\1 as select * from ${SOURCE}.\1;/" > foo
hive -f foo
echo "Updated shadow database"

18
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/monitor.sh
View File

@ -11,15 +11,15 @@ export TARGET=$2
export SHADOW=$3
export SCRIPT_PATH=$4
echo "Getting file from " $4
hdfs dfs -copyToLocal $4
export HIVE_OPTS="-hiveconf mapred.job.queue.name=analytics -hiveconf hive.spark.client.connect.timeout=120000ms -hiveconf hive.spark.client.server.connect.timeout=300000ms -hiveconf spark.executor.memory=19166291558 -hiveconf spark.yarn.executor.memoryOverhead=3225 -hiveconf spark.driver.memory=11596411699 -hiveconf spark.yarn.driver.memoryOverhead=1228"
export HADOOP_USER_NAME="oozie"
echo "Getting file from " $SCRIPT_PATH
hdfs dfs -copyToLocal $SCRIPT_PATH
echo "Creating monitor database"
cat step20-createMonitorDB.sql | sed s/SOURCE/$1/g | sed s/TARGET/$2/g1 | impala-shell -f -
echo "Impala shell finished"
#cat step20-createMonitorDB.sql | sed s/SOURCE/$1/g | sed s/TARGET/$2/g1 > foo
cat step20-createMonitorDB.sql | sed "s/TARGET/${TARGET}/g" | sed "s/SOURCE/${SOURCE}/g1" > foo
hive $HIVE_OPTS -f foo
echo "Hive shell finished"
echo "Updating shadow monitor database"
impala-shell -q "create database if not exists ${SHADOW}"
impala-shell -d ${SHADOW} -q "show tables" --delimited | sed "s/^/drop view if exists ${SHADOW}./" | sed "s/$/;/" | impala-shell -f -
impala-shell -d ${TARGET} -q "show tables" --delimited | sed "s/\(.*\)/create view ${SHADOW}.\1 as select * from ${TARGET}.\1;/" | impala-shell -f -
echo "Shadow db ready!"

20
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/observatory-post.sh
View File

@ -7,15 +7,13 @@ then
fi
export SOURCE=$1
export TARGET=$2
export SHADOW=$3
export SHADOW=$2
export HIVE_OPTS="-hiveconf mapred.job.queue.name=analytics -hiveconf hive.spark.client.connect.timeout=120000ms -hiveconf hive.spark.client.server.connect.timeout=300000ms -hiveconf spark.executor.memory=19166291558 -hiveconf spark.yarn.executor.memoryOverhead=3225 -hiveconf spark.driver.memory=11596411699 -hiveconf spark.yarn.driver.memoryOverhead=1228"
export HADOOP_USER_NAME="oozie"
impala-shell -q "invalidate metadata;"
impala-shell -d ${TARGET} -q "show tables" --delimited | sed "s/\(.*\)/compute stats ${TARGET}.\1;/" | impala-shell -f -
echo "Impala shell finished"
echo "Updating shadow observatory database"
impala-shell -q "create database if not exists ${SHADOW}"
impala-shell -d ${SHADOW} -q "show tables" --delimited | sed "s/^/drop view if exists ${SHADOW}./" | sed "s/$/;/" | impala-shell -f -
impala-shell -d ${TARGET} -q "show tables" --delimited | sed "s/\(.*\)/create view ${SHADOW}.\1 as select * from ${TARGET}.\1;/" | impala-shell -f -
echo "Shadow db ready!"
echo "Updating shadow database"
hive -e "drop database if exists ${SHADOW} cascade"
hive -e "create database if not exists ${SHADOW}"
hive $HIVE_OPTS --database ${SOURCE} -e "show tables" | grep -v WARN | sed "s/\(.*\)/create view ${SHADOW}.\1 as select * from ${SOURCE}.\1;/" > foo
hive -f foo
echo "Updated shadow database"

10
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/observatory-pre.sh
View File

@ -10,7 +10,11 @@ export SOURCE=$1
export TARGET=$2
export SHADOW=$3
export HIVE_OPTS="-hiveconf mapred.job.queue.name=analytics -hiveconf hive.spark.client.connect.timeout=120000ms -hiveconf hive.spark.client.server.connect.timeout=300000ms -hiveconf spark.executor.memory=19166291558 -hiveconf spark.yarn.executor.memoryOverhead=3225 -hiveconf spark.driver.memory=11596411699 -hiveconf spark.yarn.driver.memoryOverhead=1228"
export HADOOP_USER_NAME="oozie"
echo "Creating observatory database"
impala-shell -q "drop database if exists ${TARGET} cascade"
impala-shell -q "create database if not exists ${TARGET}"
impala-shell -d ${SOURCE} -q "show tables" --delimited | grep -iv roar | sed "s/\(.*\)/create view ${TARGET}.\1 as select * from ${SOURCE}.\1;/" | impala-shell -f -
hive -e "drop database if exists ${TARGET} cascade"
hive -e "create database if not exists ${TARGET}"
hive $HIVE_OPTS --database ${SOURCE} -e "show tables" | grep -v WARN | grep -iv roar | sed "s/\(.*\)/create view ${TARGET}.\1 as select * from ${SOURCE}.\1;/" > foo
hive -f foo

4
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/scripts/step15_5.sql
View File

@ -44,7 +44,7 @@ from (
from ${openaire_db_name}.result r lateral view explode(r.instance) instances as inst lateral view explode(inst.pid) pids as p) r
join ${stats_db_name}.result res on res.id=r.id;
create table ${stats_db_name}.result_apc as
create table ${stats_db_name}.result_apc STORED AS PARQUET as
select r.id, r.amount, r.currency
from (
select substr(r.id, 4) as id, cast(inst.processingchargeamount.value as float) as amount, inst.processingchargecurrency.value as currency
@ -52,4 +52,4 @@ from (
join ${stats_db_name}.result res on res.id=r.id
where r.amount is not null;
create view ${stats_db_name}.issn_gold_oa_dataset as select * from stats_ext.issn_gold_oa_dataset;
create view ${stats_db_name}.issn_gold_oa_dataset as select * from ${external_stats_db_name}.issn_gold_oa_dataset;

709
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/scripts/step16-createIndicatorsTables.sql
View File

File diff suppressed because it is too large Load Diff

115
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/scripts/step20-createMonitorDB.sql
View File

@ -60,87 +60,92 @@ create table TARGET.result stored as parquet as
'openorgs____::235d7f9ad18ecd7e6dc62ea4990cb9db', -- Bilkent University
'openorgs____::31f2fa9e05b49d4cf40a19c3fed8eb06', -- Saints Cyril and Methodius University of Skopje
'openorgs____::db7686f30f22cbe73a4fde872ce812a6', -- University of Milan
'openorgs____::b8b8ca674452579f3f593d9f5e557483', -- University College Cork
'openorgs____::38d7097854736583dde879d12dacafca', -- Brown University
'openorgs____::57784c9e047e826fefdb1ef816120d92', --Arts et Métiers ParisTech
'openorgs____::2530baca8a15936ba2e3297f2bce2e7e' -- University of Cape Town
))) foo;
compute stats TARGET.result;
'openorgs____::b8b8ca674452579f3f593d9f5e557483', -- University College Cork
'openorgs____::38d7097854736583dde879d12dacafca' -- Brown University
'openorgs____::57784c9e047e826fefdb1ef816120d92', --Arts et Métiers ParisTech
'openorgs____::2530baca8a15936ba2e3297f2bce2e7e', -- University of Cape Town
'openorgs____::d11f981828c485cd23d93f7f24f24db1', -- Technological University Dublin
'openorgs____::5e6bf8962665cdd040341171e5c631d8', -- Delft University of Technology
'openorgs____::846cb428d3f52a445f7275561a7beb5d' -- University of Manitoba
) )) foo;
ANALYZE TABLE TARGET.result COMPUTE STATISTICS;
create table TARGET.result_citations stored as parquet as select * from SOURCE.result_citations orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_citations;
ANALYZE TABLE TARGET.result_citations COMPUTE STATISTICS;
create table TARGET.result_references_oc stored as parquet as select * from SOURCE.result_references_oc orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_references_oc;
ANALYZE TABLE TARGET.result_references_oc COMPUTE STATISTICS;
create table TARGET.result_citations_oc stored as parquet as select * from SOURCE.result_citations_oc orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_citations_oc;
ANALYZE TABLE TARGET.result_citations_oc COMPUTE STATISTICS;
create table TARGET.result_classifications stored as parquet as select * from SOURCE.result_classifications orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_classifications;
ANALYZE TABLE TARGET.result_classifications COMPUTE STATISTICS;
create table TARGET.result_apc stored as parquet as select * from SOURCE.result_apc orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_apc;
ANALYZE TABLE TARGET.result_apc COMPUTE STATISTICS;
create table TARGET.result_concepts stored as parquet as select * from SOURCE.result_concepts orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_concepts;
ANALYZE TABLE TARGET.result_concepts COMPUTE STATISTICS;
create table TARGET.result_datasources stored as parquet as select * from SOURCE.result_datasources orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_datasources;
ANALYZE TABLE TARGET.result_datasources COMPUTE STATISTICS;
create table TARGET.result_fundercount stored as parquet as select * from SOURCE.result_fundercount orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_fundercount;
ANALYZE TABLE TARGET.result_fundercount COMPUTE STATISTICS;
create table TARGET.result_gold stored as parquet as select * from SOURCE.result_gold orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_gold;
ANALYZE TABLE TARGET.result_gold COMPUTE STATISTICS;
create table TARGET.result_greenoa stored as parquet as select * from SOURCE.result_greenoa orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_greenoa;
ANALYZE TABLE TARGET.result_greenoa COMPUTE STATISTICS;
create table TARGET.result_languages stored as parquet as select * from SOURCE.result_languages orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_languages;
ANALYZE TABLE TARGET.result_languages COMPUTE STATISTICS;
create table TARGET.result_licenses stored as parquet as select * from SOURCE.result_licenses orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_licenses;
ANALYZE TABLE TARGET.result_licenses COMPUTE STATISTICS;
create table TARGET.licenses_normalized STORED AS PARQUET as select * from SOURCE.licenses_normalized;
ANALYZE TABLE TARGET.licenses_normalized COMPUTE STATISTICS;
create table TARGET.result_oids stored as parquet as select * from SOURCE.result_oids orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_oids;
ANALYZE TABLE TARGET.result_oids COMPUTE STATISTICS;
create table TARGET.result_organization stored as parquet as select * from SOURCE.result_organization orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_organization;
ANALYZE TABLE TARGET.result_organization COMPUTE STATISTICS;
create table TARGET.result_peerreviewed stored as parquet as select * from SOURCE.result_peerreviewed orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_peerreviewed;
ANALYZE TABLE TARGET.result_peerreviewed COMPUTE STATISTICS;
create table TARGET.result_pids stored as parquet as select * from SOURCE.result_pids orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_pids;
ANALYZE TABLE TARGET.result_pids COMPUTE STATISTICS;
create table TARGET.result_projectcount stored as parquet as select * from SOURCE.result_projectcount orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_projectcount;
ANALYZE TABLE TARGET.result_projectcount COMPUTE STATISTICS;
create table TARGET.result_projects stored as parquet as select * from SOURCE.result_projects orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_projects;
ANALYZE TABLE TARGET.result_projects COMPUTE STATISTICS;
create table TARGET.result_refereed stored as parquet as select * from SOURCE.result_refereed orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_refereed;
ANALYZE TABLE TARGET.result_refereed COMPUTE STATISTICS;
create table TARGET.result_sources stored as parquet as select * from SOURCE.result_sources orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_sources;
ANALYZE TABLE TARGET.result_sources COMPUTE STATISTICS;
create table TARGET.result_topics stored as parquet as select * from SOURCE.result_topics orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_topics;
ANALYZE TABLE TARGET.result_topics COMPUTE STATISTICS;
create table TARGET.result_fos stored as parquet as select * from SOURCE.result_fos orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.result_fos;
ANALYZE TABLE TARGET.result_fos COMPUTE STATISTICS;
create view TARGET.foo1 as select * from SOURCE.result_result rr where rr.source in (select id from TARGET.result);
create view TARGET.foo2 as select * from SOURCE.result_result rr where rr.target in (select id from TARGET.result);
create table TARGET.result_result STORED AS PARQUET as select distinct * from (select * from TARGET.foo1 union all select * from TARGET.foo2) foufou;
drop view TARGET.foo1;
drop view TARGET.foo2;
compute stats TARGET.result_result;
ANALYZE TABLE TARGET.result_result COMPUTE STATISTICS;
-- datasources
create view if not exists TARGET.datasource as select * from SOURCE.datasource;
@ -149,7 +154,7 @@ create view if not exists TARGET.datasource_organizations as select * from SOURC
create view if not exists TARGET.datasource_sources as select * from SOURCE.datasource_sources;
create table TARGET.datasource_results stored as parquet as select id as result, datasource as id from TARGET.result_datasources;
compute stats TARGET.datasource_results;
ANALYZE TABLE TARGET.datasource_results COMPUTE STATISTICS;
-- organizations
create view if not exists TARGET.organization as select * from SOURCE.organization;
@ -164,30 +169,31 @@ create view if not exists TARGET.project_oids as select * from SOURCE.project_oi
create view if not exists TARGET.project_organizations as select * from SOURCE.project_organizations;
create view if not exists TARGET.project_resultcount as select * from SOURCE.project_resultcount;
create view if not exists TARGET.project_classification as select * from SOURCE.project_classification;
create view if not exists TARGET.project_organization_contribution as select * from SOURCE.project_organization_contribution;
create table TARGET.project_results stored as parquet as select id as result, project as id from TARGET.result_projects;
compute stats TARGET.project_results;
ANALYZE TABLE TARGET.project_results COMPUTE STATISTICS;
-- indicators
-- Sprint 1 ----
create table TARGET.indi_pub_green_oa stored as parquet as select * from SOURCE.indi_pub_green_oa orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.indi_pub_green_oa;
ANALYZE TABLE TARGET.indi_pub_green_oa COMPUTE STATISTICS;
create table TARGET.indi_pub_grey_lit stored as parquet as select * from SOURCE.indi_pub_grey_lit orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.indi_pub_grey_lit;
ANALYZE TABLE TARGET.indi_pub_grey_lit COMPUTE STATISTICS;
create table TARGET.indi_pub_doi_from_crossref stored as parquet as select * from SOURCE.indi_pub_doi_from_crossref orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.indi_pub_doi_from_crossref;
ANALYZE TABLE TARGET.indi_pub_doi_from_crossref COMPUTE STATISTICS;
-- Sprint 2 ----
create table TARGET.indi_result_has_cc_licence stored as parquet as select * from SOURCE.indi_result_has_cc_licence orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.indi_result_has_cc_licence;
ANALYZE TABLE TARGET.indi_result_has_cc_licence COMPUTE STATISTICS;
create table TARGET.indi_result_has_cc_licence_url stored as parquet as select * from SOURCE.indi_result_has_cc_licence_url orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.indi_result_has_cc_licence_url;
ANALYZE TABLE TARGET.indi_result_has_cc_licence_url COMPUTE STATISTICS;
create table TARGET.indi_pub_has_abstract stored as parquet as select * from SOURCE.indi_pub_has_abstract orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.indi_pub_has_abstract;
ANALYZE TABLE TARGET.indi_pub_has_abstract COMPUTE STATISTICS;
create table TARGET.indi_result_with_orcid stored as parquet as select * from SOURCE.indi_result_with_orcid orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.indi_result_with_orcid;
ANALYZE TABLE TARGET.indi_result_with_orcid COMPUTE STATISTICS;
---- Sprint 3 ----
create table TARGET.indi_funded_result_with_fundref stored as parquet as select * from SOURCE.indi_funded_result_with_fundref orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.indi_funded_result_with_fundref;
ANALYZE TABLE TARGET.indi_funded_result_with_fundref COMPUTE STATISTICS;
create view TARGET.indi_result_org_collab as select * from SOURCE.indi_result_org_collab;
create view TARGET.indi_result_org_country_collab as select * from SOURCE.indi_result_org_country_collab;
create view TARGET.indi_project_collab_org as select * from SOURCE.indi_project_collab_org;
@ -196,30 +202,30 @@ create view TARGET.indi_funder_country_collab as select * from SOURCE.indi_funde
create view TARGET.indi_result_country_collab as select * from SOURCE.indi_result_country_collab;
---- Sprint 4 ----
create table TARGET.indi_pub_diamond stored as parquet as select * from SOURCE.indi_pub_diamond orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.indi_pub_diamond;
ANALYZE TABLE TARGET.indi_pub_diamond COMPUTE STATISTICS;
create table TARGET.indi_pub_in_transformative stored as parquet as select * from SOURCE.indi_pub_in_transformative orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.indi_pub_in_transformative;
ANALYZE TABLE TARGET.indi_pub_in_transformative COMPUTE STATISTICS;
create table TARGET.indi_pub_closed_other_open stored as parquet as select * from SOURCE.indi_pub_closed_other_open orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.indi_pub_closed_other_open;
ANALYZE TABLE TARGET.indi_pub_closed_other_open COMPUTE STATISTICS;
---- Sprint 5 ----
create table TARGET.indi_result_no_of_copies stored as parquet as select * from SOURCE.indi_result_no_of_copies orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.indi_result_no_of_copies;
ANALYZE TABLE TARGET.indi_result_no_of_copies COMPUTE STATISTICS;
---- Sprint 6 ----
create table TARGET.indi_pub_hybrid_oa_with_cc stored as parquet as select * from SOURCE.indi_pub_hybrid_oa_with_cc orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.indi_pub_hybrid_oa_with_cc;
ANALYZE TABLE TARGET.indi_pub_hybrid_oa_with_cc COMPUTE STATISTICS;
create table TARGET.indi_pub_downloads stored as parquet as select * from SOURCE.indi_pub_downloads orig where exists (select 1 from TARGET.result r where r.id=orig.result_id);
compute stats TARGET.indi_pub_downloads;
ANALYZE TABLE TARGET.indi_pub_downloads COMPUTE STATISTICS;
create table TARGET.indi_pub_downloads_datasource stored as parquet as select * from SOURCE.indi_pub_downloads_datasource orig where exists (select 1 from TARGET.result r where r.id=orig.result_id);
compute stats TARGET.indi_pub_downloads_datasource;
ANALYZE TABLE TARGET.indi_pub_downloads_datasource COMPUTE STATISTICS;
create table TARGET.indi_pub_downloads_year stored as parquet as select * from SOURCE.indi_pub_downloads_year orig where exists (select 1 from TARGET.result r where r.id=orig.result_id);
compute stats TARGET.indi_pub_downloads_year;
ANALYZE TABLE TARGET.indi_pub_downloads_year COMPUTE STATISTICS;
create table TARGET.indi_pub_downloads_datasource_year stored as parquet as select * from SOURCE.indi_pub_downloads_datasource_year orig where exists (select 1 from TARGET.result r where r.id=orig.result_id);
compute stats TARGET.indi_pub_downloads_datasource_year;
ANALYZE TABLE TARGET.indi_pub_downloads_datasource_year COMPUTE STATISTICS;
---- Sprint 7 ----
create table TARGET.indi_pub_gold_oa stored as parquet as select * from SOURCE.indi_pub_gold_oa orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.indi_pub_gold_oa;
ANALYZE TABLE TARGET.indi_pub_gold_oa COMPUTE STATISTICS;
create table TARGET.indi_pub_hybrid stored as parquet as select * from SOURCE.indi_pub_hybrid orig where exists (select 1 from TARGET.result r where r.id=orig.id);
compute stats TARGET.indi_pub_hybrid;
ANALYZE TABLE TARGET.indi_pub_hybrid COMPUTE STATISTICS;
create view TARGET.indi_org_fairness as select * from SOURCE.indi_org_fairness;
create view TARGET.indi_org_fairness_pub_pr as select * from SOURCE.indi_org_fairness_pub_pr;
create view TARGET.indi_org_fairness_pub_year as select * from SOURCE.indi_org_fairness_pub_year;
@ -230,11 +236,8 @@ create view TARGET.indi_org_findable as select * from SOURCE.indi_org_findable;
create view TARGET.indi_org_openess as select * from SOURCE.indi_org_openess;
create view TARGET.indi_org_openess_year as select * from SOURCE.indi_org_openess_year;
create table TARGET.indi_pub_has_preprint stored as parquet as select * from SOURCE.indi_pub_has_preprint orig where exists (select 1 from TARGET.result r where r.id=orig.id);
ANALYZE TABLE TARGET.indi_pub_has_preprint COMPUTE STATISTICS;
create table TARGET.indi_pub_in_subscribed stored as parquet as select * from SOURCE.indi_pub_in_subscribed orig where exists (select 1 from TARGET.result r where r.id=orig.id);
ANALYZE TABLE TARGET.indi_pub_in_subscribed COMPUTE STATISTICS;
create table TARGET.indi_result_with_pid stored as parquet as select * from SOURCE.indi_result_with_pid orig where exists (select 1 from TARGET.result r where r.id=orig.id);
--create table TARGET.indi_datasets_gold_oa stored as parquet as select * from SOURCE.indi_datasets_gold_oa orig where exists (select 1 from TARGET.result r where r.id=orig.id);
--compute stats TARGET.indi_datasets_gold_oa;
--create table TARGET.indi_software_gold_oa stored as parquet as select * from SOURCE.indi_software_gold_oa orig where exists (select 1 from TARGET.result r where r.id=orig.id);
--compute stats TARGET.indi_software_gold_oa;
ANALYZE TABLE TARGET.indi_result_with_pid COMPUTE STATISTICS;

40
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/scripts/step21-createObservatoryDB.sql
View File

@ -8,6 +8,8 @@ from ${stats_db_name}.result r
group by rl.id
) rln on rln.id=r.id;
ANALYZE TABLE ${observatory_db_name}.result_cc_licence COMPUTE STATISTICS;
create table ${observatory_db_name}.result_affiliated_country stored as parquet as
select
count(distinct r.id) as total,
@ -37,6 +39,8 @@ group by r.green, r.gold, case when rl.type is not null then true else false end
case when r.access_mode in ('Open Access', 'Open Source') then true else false end, r.peer_reviewed, r.type, abstract,
cc_licence, r.authors > 1, rpc.count > 1, rfc.count > 1, c.code, c.name;
ANALYZE TABLE ${observatory_db_name}.result_affiliated_country COMPUTE STATISTICS;
create table ${observatory_db_name}.result_affiliated_year stored as parquet as
select
count(distinct r.id) as total,
@ -66,6 +70,8 @@ group by r.green, r.gold, case when rl.type is not null then true else false end
case when r.access_mode in ('Open Access', 'Open Source') then true else false end, r.peer_reviewed, r.type, abstract,
cc_licence, r.authors > 1, rpc.count > 1, rfc.count > 1, r.year;
ANALYZE TABLE ${observatory_db_name}.result_affiliated_year COMPUTE STATISTICS;
create table ${observatory_db_name}.result_affiliated_year_country stored as parquet as
select
count(distinct r.id) as total,
@ -95,6 +101,8 @@ group by r.green, r.gold, case when rl.type is not null then true else false end
case when r.access_mode in ('Open Access', 'Open Source') then true else false end, r.peer_reviewed, r.type, abstract,
cc_licence, r.authors > 1, rpc.count > 1, rfc.count > 1, r.year, c.code, c.name;
ANALYZE TABLE ${observatory_db_name}.result_affiliated_year_country COMPUTE STATISTICS;
create table ${observatory_db_name}.result_affiliated_datasource stored as parquet as
select
count(distinct r.id) as total,
@ -126,6 +134,8 @@ group by r.green, r.gold, case when rl.type is not null then true else false end
case when r.access_mode in ('Open Access', 'Open Source') then true else false end, r.peer_reviewed, r.type, abstract,
cc_licence, r.authors > 1, rpc.count > 1, rfc.count > 1, d.name;
ANALYZE TABLE ${observatory_db_name}.result_affiliated_datasource COMPUTE STATISTICS;
create table ${observatory_db_name}.result_affiliated_datasource_country stored as parquet as
select
count(distinct r.id) as total,
@ -157,6 +167,8 @@ group by r.green, r.gold, case when rl.type is not null then true else false end
case when r.access_mode in ('Open Access', 'Open Source') then true else false end, r.peer_reviewed, r.type, abstract,
cc_licence, r.authors > 1, rpc.count > 1, rfc.count > 1, d.name, c.code, c.name;
ANALYZE TABLE ${observatory_db_name}.result_affiliated_datasource_country COMPUTE STATISTICS;
create table ${observatory_db_name}.result_affiliated_organization stored as parquet as
select
count(distinct r.id) as total,
@ -186,6 +198,8 @@ group by r.green, r.gold, case when rl.type is not null then true else false end
case when r.access_mode in ('Open Access', 'Open Source') then true else false end, r.peer_reviewed, r.type, abstract,
cc_licence, r.authors > 1, rpc.count > 1, rfc.count > 1, o.name;
ANALYZE TABLE ${observatory_db_name}.result_affiliated_organization COMPUTE STATISTICS;
create table ${observatory_db_name}.result_affiliated_organization_country stored as parquet as
select
count(distinct r.id) as total,
@ -215,6 +229,8 @@ group by r.green, r.gold, case when rl.type is not null then true else false end
case when r.access_mode in ('Open Access', 'Open Source') then true else false end, r.peer_reviewed, r.type, abstract,
cc_licence, r.authors > 1, rpc.count > 1, rfc.count > 1, o.name, c.code, c.name;
ANALYZE TABLE ${observatory_db_name}.result_affiliated_organization_country COMPUTE STATISTICS;
create table ${observatory_db_name}.result_affiliated_funder stored as parquet as
select
count(distinct r.id) as total,
@ -246,6 +262,8 @@ group by r.green, r.gold, case when rl.type is not null then true else false end
case when r.access_mode in ('Open Access', 'Open Source') then true else false end, r.peer_reviewed, r.type, abstract,
cc_licence, r.authors > 1, rpc.count > 1, rfc.count > 1, p.funder;
ANALYZE TABLE ${observatory_db_name}.result_affiliated_funder COMPUTE STATISTICS;
create table ${observatory_db_name}.result_affiliated_funder_country stored as parquet as
select
count(distinct r.id) as total,
@ -277,6 +295,8 @@ group by r.green, r.gold, case when rl.type is not null then true else false end
case when r.access_mode in ('Open Access', 'Open Source') then true else false end, r.peer_reviewed, r.type, abstract,
cc_licence, r.authors > 1, rpc.count > 1, rfc.count > 1, p.funder, c.code, c.name;
ANALYZE TABLE ${observatory_db_name}.result_affiliated_funder_country COMPUTE STATISTICS;
create table ${observatory_db_name}.result_deposited_country stored as parquet as
select
count(distinct r.id) as total,
@ -308,6 +328,8 @@ group by r.green, r.gold, case when rl.type is not null then true else false end
case when r.access_mode in ('Open Access', 'Open Source') then true else false end, r.peer_reviewed, r.type, abstract,
cc_licence, r.authors > 1, rpc.count > 1, rfc.count > 1, c.code, c.name;
ANALYZE TABLE ${observatory_db_name}.result_deposited_country COMPUTE STATISTICS;
create table ${observatory_db_name}.result_deposited_year stored as parquet as
select
count(distinct r.id) as total,
@ -339,6 +361,8 @@ group by r.green, r.gold, case when rl.type is not null then true else false end
case when r.access_mode in ('Open Access', 'Open Source') then true else false end, r.peer_reviewed, r.type, abstract,
cc_licence, r.authors > 1, rpc.count > 1, rfc.count > 1, r.year;
ANALYZE TABLE ${observatory_db_name}.result_deposited_year COMPUTE STATISTICS;
create table ${observatory_db_name}.result_deposited_year_country stored as parquet as
select
count(distinct r.id) as total,
@ -370,6 +394,8 @@ group by r.green, r.gold, case when rl.type is not null then true else false end
case when r.access_mode in ('Open Access', 'Open Source') then true else false end, r.peer_reviewed, r.type, abstract,
cc_licence, r.authors > 1, rpc.count > 1, rfc.count > 1, r.year, c.code, c.name;
ANALYZE TABLE ${observatory_db_name}.result_deposited_year_country COMPUTE STATISTICS;
create table ${observatory_db_name}.result_deposited_datasource stored as parquet as
select
count(distinct r.id) as total,
@ -401,6 +427,8 @@ group by r.green, r.gold, case when rl.type is not null then true else false end
case when r.access_mode in ('Open Access', 'Open Source') then true else false end, r.peer_reviewed, r.type, abstract,
cc_licence, r.authors > 1, rpc.count > 1, rfc.count > 1, d.name;
ANALYZE TABLE ${observatory_db_name}.result_deposited_datasource COMPUTE STATISTICS;
create table ${observatory_db_name}.result_deposited_datasource_country stored as parquet as
select
count(distinct r.id) as total,
@ -432,6 +460,8 @@ group by r.green, r.gold, case when rl.type is not null then true else false end
case when r.access_mode in ('Open Access', 'Open Source') then true else false end, r.peer_reviewed, r.type, abstract,
cc_licence, r.authors > 1, rpc.count > 1, rfc.count > 1, d.name, c.code, c.name;
ANALYZE TABLE ${observatory_db_name}.result_deposited_datasource_country COMPUTE STATISTICS;
create table ${observatory_db_name}.result_deposited_organization stored as parquet as
select
count(distinct r.id) as total,
@ -463,6 +493,8 @@ group by r.green, r.gold, case when rl.type is not null then true else false end
case when r.access_mode in ('Open Access', 'Open Source') then true else false end, r.peer_reviewed, r.type, abstract,
cc_licence, r.authors > 1, rpc.count > 1, rfc.count > 1, o.name;
ANALYZE TABLE ${observatory_db_name}.result_deposited_organization COMPUTE STATISTICS;
create table ${observatory_db_name}.result_deposited_organization_country stored as parquet as
select
count(distinct r.id) as total,
@ -494,6 +526,8 @@ group by r.green, r.gold, case when rl.type is not null then true else false end
case when r.access_mode in ('Open Access', 'Open Source') then true else false end, r.peer_reviewed, r.type, abstract,
cc_licence, r.authors > 1, rpc.count > 1, rfc.count > 1, o.name, c.code, c.name;
ANALYZE TABLE ${observatory_db_name}.result_deposited_organization_country COMPUTE STATISTICS;
create table ${observatory_db_name}.result_deposited_funder stored as parquet as
select
count(distinct r.id) as total,
@ -527,6 +561,8 @@ group by r.green, r.gold, case when rl.type is not null then true else false end
case when r.access_mode in ('Open Access', 'Open Source') then true else false end, r.peer_reviewed, r.type, abstract,
cc_licence, r.authors > 1, rpc.count > 1, rfc.count > 1, p.funder;
ANALYZE TABLE ${observatory_db_name}.result_deposited_funder COMPUTE STATISTICS;
create table ${observatory_db_name}.result_deposited_funder_country stored as parquet as
select
count(distinct r.id) as total,
@ -558,4 +594,6 @@ from ${stats_db_name}.result r
left outer join ${stats_db_name}.result_fundercount rfc on rfc.id=r.id
group by r.green, r.gold, case when rl.type is not null then true else false end, case when pids.pid is not null then true else false end,
case when r.access_mode in ('Open Access', 'Open Source') then true else false end, r.peer_reviewed, r.type, abstract,
cc_licence, r.authors > 1, rpc.count > 1, rfc.count > 1, p.funder, c.code, c.name;
cc_licence, r.authors > 1, rpc.count > 1, rfc.count > 1, p.funder, c.code, c.name;
ANALYZE TABLE ${observatory_db_name}.result_deposited_funder_country COMPUTE STATISTICS;

10
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/scripts/step6.sql
View File

@ -84,4 +84,12 @@ create table ${stats_db_name}.funder STORED AS PARQUET as
select distinct xpath_string(fund, '//funder/id') as id,
xpath_string(fund, '//funder/name') as name,
xpath_string(fund, '//funder/shortname') as shortname
from ${openaire_db_name}.project p lateral view explode(p.fundingtree.value) fundingtree as fund;
from ${openaire_db_name}.project p lateral view explode(p.fundingtree.value) fundingtree as fund;
CREATE TABLE ${stats_db_name}.project_organization_contribution STORED AS PARQUET AS
SELECT distinct substr(r.source, 4) AS project, substr(r.target, 4) AS organization,
properties[0].value contribution, properties[1].value currency
from ${openaire_db_name}.relation r
LATERAL VIEW explode (r.properties) properties
where properties[0].key='contribution' and r.reltype = 'projectOrganization' and r.source like '40|%'
and properties[0].value>0.0 and r.datainfo.deletedbyinference = false and r.datainfo.invisible=false;

20
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/scripts/step8.sql
View File

@ -48,12 +48,10 @@ WHERE d1.datainfo.deletedbyinference = FALSE and d1.datainfo.invisible=false;
-- Updating temporary table with everything that is not based on results -> This is done with the following "dual" table.
-- Creating a temporary dual table that will be removed after the following insert
CREATE TABLE ${stats_db_name}.dual
(
dummy CHAR(1)
);
INSERT INTO ${stats_db_name}.dual
VALUES ('X');
CREATE TABLE ${stats_db_name}.dual ( dummy CHAR(1));
INSERT INTO ${stats_db_name}.dual VALUES ('X');
INSERT INTO ${stats_db_name}.datasource_tmp (`id`, `name`, `type`, `dateofvalidation`, `yearofvalidation`, `harvested`,
`piwik_id`, `latitude`, `longitude`, `websiteurl`, `compatibility`, `issn_printed`, `issn_online`)
SELECT 'other',
@ -73,12 +71,8 @@ FROM ${stats_db_name}.dual
WHERE 'other' not in (SELECT id FROM ${stats_db_name}.datasource_tmp WHERE name = 'Unknown Repository');
DROP TABLE ${stats_db_name}.dual;
UPDATE ${stats_db_name}.datasource_tmp
SET name='Other'
WHERE name = 'Unknown Repository';
UPDATE ${stats_db_name}.datasource_tmp
SET yearofvalidation=null
WHERE yearofvalidation = '-1';
UPDATE ${stats_db_name}.datasource_tmp SET name='Other' WHERE name = 'Unknown Repository';
UPDATE ${stats_db_name}.datasource_tmp SET yearofvalidation=null WHERE yearofvalidation = '-1';
CREATE TABLE ${stats_db_name}.datasource_languages STORED AS PARQUET AS
SELECT substr(d.id, 4) AS id, langs.languages AS language
@ -104,4 +98,4 @@ where d.datainfo.deletedbyinference = false and d.datainfo.invisible=false;
CREATE OR REPLACE VIEW ${stats_db_name}.datasource_results AS
SELECT datasource AS id, id AS result
FROM ${stats_db_name}.result_datasources;
FROM ${stats_db_name}.result_datasources;

76
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/workflow.xml
View File

@ -1,4 +1,4 @@
<workflow-app name="Graph Stats" xmlns="uri:oozie:workflow:0.5">
<workflow-app name="Graph Stats Hive" xmlns="uri:oozie:workflow:0.5">
<parameters>
<property>
<name>stats_db_name</name>
@ -10,7 +10,6 @@
</property>
<property>
<name>external_stats_db_name</name>
<value>stats_ext</value>
<description>the external stats that should be added since they are not included in the graph database</description>
</property>
<property>
@ -53,6 +52,10 @@
<name>context_api_url</name>
<description>the base url of the context api (https://services.openaire.eu/openaire)</description>
</property>
<property>
<name>hadoop_user_name</name>
<description>user name of the wf owner</description>
</property>
</parameters>
<global>
@ -67,11 +70,14 @@
<name>hive.txn.timeout</name>
<value>${hive_timeout}</value>
</property>
<property>
<name>mapred.job.queue.name</name>
<value>analytics</value>
</property>
</configuration>
</global>
<start to="resume_from"/>
<decision name="resume_from">
<switch>
<case to="Step1">${wf:conf('resumeFrom') eq 'Step1'}</case>
@ -99,12 +105,13 @@
<case to="step21-createObservatoryDB-pre">${wf:conf('resumeFrom') eq 'step21-createObservatoryDB-pre'}</case>
<case to="step21-createObservatoryDB">${wf:conf('resumeFrom') eq 'step21-createObservatoryDB'}</case>
<case to="step21-createObservatoryDB-post">${wf:conf('resumeFrom') eq 'step21-createObservatoryDB-post'}</case>
<case to="Step22">${wf:conf('resumeFrom') eq 'Step22'}</case>
<case to="step22-copyDataToImpalaCluster">${wf:conf('resumeFrom') eq 'step22-copyDataToImpalaCluster'}</case>
<case to="step23-finalizeImpalaCluster">${wf:conf('resumeFrom') eq 'step23-finalizeImpalaCluster'}</case>
<case to="Step24-updateCache">${wf:conf('resumeFrom') eq 'Step24-updateCache'}</case>
<default to="Step1"/>
</switch>
</decision>
<kill name="Kill">
<message>Action failed, error message[${wf:errorMessage(wf:lastErrorNode())}]</message>
</kill>
@ -282,6 +289,7 @@
<script>scripts/step15_5.sql</script>
<param>stats_db_name=${stats_db_name}</param>
<param>openaire_db_name=${openaire_db_name}</param>
<param>external_stats_db_name=${external_stats_db_name}</param>
</hive2>
<ok to="Contexts"/>
<error to="Kill"/>
@ -306,6 +314,7 @@
<name-node>${nameNode}</name-node>
<exec>indicators.sh</exec>
<argument>${stats_db_name}</argument>
<argument>${external_stats_db_name}</argument>
<argument>${wf:appPath()}/scripts/step16-createIndicatorsTables.sql</argument>
<file>indicators.sh</file>
</shell>
@ -334,7 +343,7 @@
<ok to="Step19-finalize"/>
<error to="Kill"/>
</action>
<action name="Step19-finalize">
<shell xmlns="uri:oozie:shell-action:0.1">
<job-tracker>${jobTracker}</job-tracker>
@ -359,6 +368,19 @@
<argument>${wf:appPath()}/scripts/step20-createMonitorDB.sql</argument>
<file>monitor.sh</file>
</shell>
<ok to="step20-createMonitorDB-post"/>
<error to="Kill"/>
</action>
<action name="step20-createMonitorDB-post">
<shell xmlns="uri:oozie:shell-action:0.1">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<exec>monitor-post.sh</exec>
<argument>${monitor_db_name}</argument>
<argument>${monitor_db_shadow_name}</argument>
<file>monitor-post.sh</file>
</shell>
<ok to="step21-createObservatoryDB-pre"/>
<error to="Kill"/>
</action>
@ -393,16 +415,50 @@
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<exec>observatory-post.sh</exec>
<argument>${stats_db_name}</argument>
<argument>${observatory_db_name}</argument>
<argument>${observatory_db_shadow_name}</argument>
<file>observatory-post.sh</file>
</shell>
<ok to="Step22"/>
<ok to="step22-copyDataToImpalaCluster"/>
<error to="Kill"/>
</action>
<action name="Step22">
<action name="step22-copyDataToImpalaCluster">
<shell xmlns="uri:oozie:shell-action:0.1">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<exec>copyDataToImpalaCluster.sh</exec>
<!-- <env-var>HADOOP_USER_NAME=${wf:user()}</env-var>-->
<!-- <argument>${external_stats_db_name}</argument>-->
<argument>${stats_db_name}</argument>
<argument>${monitor_db_name}</argument>
<argument>${observatory_db_name}</argument>
<argument>${external_stats_db_name}</argument>
<argument>${hadoop_user_name}</argument>
<file>copyDataToImpalaCluster.sh</file>
</shell>
<ok to="step23-finalizeImpalaCluster"/>
<error to="Kill"/>
</action>
<action name="step23-finalizeImpalaCluster">
<shell xmlns="uri:oozie:shell-action:0.1">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
<exec>finalizeImpalaCluster.sh</exec>
<argument>${stats_db_name}</argument>
<argument>${stats_db_shadow_name}</argument>
<argument>${monitor_db_name}</argument>
<argument>${monitor_db_shadow_name}</argument>
<argument>${observatory_db_name}</argument>
<argument>${observatory_db_shadow_name}</argument>
<file>finalizeImpalaCluster.sh</file>
</shell>
<ok to="Step24-updateCache"/>
<error to="Kill"/>
</action>
<action name="Step24-updateCache">
<shell xmlns="uri:oozie:shell-action:0.1">
<job-tracker>${jobTracker}</job-tracker>
<name-node>${nameNode}</name-node>
@ -415,4 +471,4 @@
</action>
<end name="End"/>
</workflow-app>
</workflow-app>