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

using distinct apcs per publication to avoid huge sums

This commit is contained in:
Antonis Lempesis 2024-03-07 02:07:59 +02:00
parent 2e4cab026c
commit 316d585c8a
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/scripts/step15_5.sql
View File

@ -56,7 +56,7 @@ join ${stats_db_name}.result res on res.id=r.id;
DROP TABLE IF EXISTS ${stats_db_name}.result_apc purge;
create table if not exists ${stats_db_name}.result_apc STORED AS PARQUET as
select r.id, r.amount, r.currency
select distinct 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
from ${openaire_db_name}.result r lateral view explode(r.instance) instances as inst) r