D-Net
/
dnet-hadoop
14
5
Fork 11
Code Issues 14 Pull Requests 13 Projects 2 Releases 2 Wiki Activity

Improve performance of creating the "result_fos" table, by using a temp-table to cache data, which is requested multiple times.

This commit is contained in:
Lampros Smyrnaios 2024-07-03 20:24:36 +03:00
parent ce0aee21cc
commit e9686365a2
1 changed files with 10 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/scripts/step7.sql
View File

@ -125,14 +125,20 @@ UNION ALL
SELECT *
FROM ${stats_db_name}.otherresearchproduct_topics; /*EOS*/
DROP TABLE IF EXISTS ${stats_db_name}.result_fos_base_tmp purge; /*EOS*/
create table ${stats_db_name}.result_fos_base_tmp stored as parquet as
select /*+ COALESCE(100) */ id, topic from ${stats_db_name}.result_topics where type='Fields of Science and Technology classification'; /*EOS*/
DROP TABLE IF EXISTS ${stats_db_name}.result_fos purge; /*EOS*/
create table ${stats_db_name}.result_fos stored as parquet as
with
lvl1 as (select id, topic from ${stats_db_name}.result_topics where topic like '__ %' and type='Fields of Science and Technology classification'),
lvl2 as (select id, topic from ${stats_db_name}.result_topics where topic like '____ %' and type='Fields of Science and Technology classification'),
lvl3 as (select id, topic from ${stats_db_name}.result_topics where topic like '______ %' and type='Fields of Science and Technology classification'),
lvl4 as (select id, topic from ${stats_db_name}.result_topics where topic like '________ %' and type='Fields of Science and Technology classification')
lvl1 as (select * from ${stats_db_name}.result_fos_base_tmp where topic like '__ %'),
lvl2 as (select * from ${stats_db_name}.result_fos_base_tmp where topic like '____ %'),
lvl3 as (select * from ${stats_db_name}.result_fos_base_tmp where topic like '______ %'),
lvl4 as (select * from ${stats_db_name}.result_fos_base_tmp where topic like '________ %')
select /*+ COALESCE(100) */ lvl1.id, lvl1.topic as lvl1, lvl2.topic as lvl2, lvl3.topic as lvl3, lvl4.topic as lvl4
from lvl1
join lvl2 on lvl1.id=lvl2.id and substr(lvl2.topic, 1, 2)=substr(lvl1.topic, 1, 2)