[DOIBOOST Process] fix filtering to filter results with non null id

2021-11-25 12:10:45 +01:00 · 2021-11-25 12:10:45 +01:00 · 5fd0e610bf
parent a7cf277d98
commit 5fd0e610bf
1 changed files with 1 additions and 1 deletions
--- a/dhp-workflows/dhp-doiboost/src/main/java/eu/dnetlib/doiboost/mag/SparkProcessMAG.scala
+++ b/dhp-workflows/dhp-doiboost/src/main/java/eu/dnetlib/doiboost/mag/SparkProcessMAG.scala
@ -146,7 +146,7 @@ object SparkProcessMAG {
      .save(s"$workingPath/mag_publication")

    spark.read.load(s"$workingPath/mag_publication").as[Publication]
-      .filter(p => p.getId == null)
+      .filter(p => p.getId != null)
      .groupByKey(p => p.getId)
      .reduceGroups((a:Publication, b:Publication) => ConversionUtil.mergePublication(a,b))
      .map(_._2)