D-Net
/
dnet-hadoop
13
4
Fork 10
Code Issues 14 Pull Requests 8 Projects 2 Releases 1 Wiki Activity

Patch the identifiers (source/target) in the relations, refinements #131

Merged
claudio.atzori merged 6 commits from fct_project_id_replacement into master 2021-07-30 11:07:54 +02:00
Conversation 0 Commits 6 Files Changed 5 +4 -3
1 changed files with 4 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files
Showing only changes of commit 9bc4fd3b69 - Show all commits

7
dhp-workflows/dhp-graph-mapper/src/main/java/eu/dnetlib/dhp/oa/graph/raw/PatchRelationsApplication.java
View File

@ -82,7 +82,7 @@ public class PatchRelationsApplication {
log.info("relations: {}", rels.count());
log.info("idMapping: {}", idMapping.count());
rels
Dataset<Relation> fj = rels
.joinWith(idMapping, rels.col("source").equalTo(idMapping.col("oldId")), "left")
.map((MapFunction<Tuple2<Relation, RelationIdMapping>, Relation>) t -> {
final Relation r = t._1();
@ -90,8 +90,9 @@ public class PatchRelationsApplication {
.map(RelationIdMapping::getNewId)
.ifPresent(r::setSource);
return r;
}, Encoders.bean(Relation.class))
.joinWith(idMapping, rels.col("target").equalTo(idMapping.col("oldId")), "left")
}, Encoders.bean(Relation.class));
fj.joinWith(idMapping, fj.col("target").equalTo(idMapping.col("oldId")), "left")
.map((MapFunction<Tuple2<Relation, RelationIdMapping>, Relation>) t -> {
final Relation r = t._1();
Optional.ofNullable(t._2())