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

Changes from beta stats wf to prod #264

Merged
claudio.atzori merged 8 commits from antonis.lempesis/dnet-hadoop:beta into master 2022-12-07 15:56:22 +01:00
Conversation 2 Commits 8 Files Changed 7 +5 -5
1 changed files with 5 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files
Showing only changes of commit 2ebb1459a9 - Show all commits

10
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/scripts/step16-createIndicatorsTables.sql
View File

@ -454,16 +454,16 @@ FROM publication_datasources pd
compute stats indi_pub_hybrid_oa_with_cc; compute stats indi_pub_hybrid_oa_with_cc;
create table indi_pub_downloads stored as parquet as create table indi_pub_downloads stored as parquet as
SELECT result_id, sum(downloads) no_dowloads from openaire_prod_usage_stats.usage_stats SELECT result_id, sum(downloads) no_downloads from openaire_prod_usage_stats.usage_stats
join publication on result_id=id join publication on result_id=id
where downloads>0 where downloads>0
GROUP BY result_id GROUP BY result_id
order by no_dowloads desc; order by no_downloads desc;
compute stats indi_pub_downloads; compute stats indi_pub_downloads;
create table indi_pub_downloads_datasource stored as parquet as create table indi_pub_downloads_datasource stored as parquet as
SELECT result_id, repository_id, sum(downloads) no_dowloads from openaire_prod_usage_stats.usage_stats SELECT result_id, repository_id, sum(downloads) no_downloads from openaire_prod_usage_stats.usage_stats
join publication on result_id=id join publication on result_id=id
where downloads>0 where downloads>0
GROUP BY result_id, repository_id GROUP BY result_id, repository_id
@ -472,7 +472,7 @@ order by result_id;
compute stats indi_pub_downloads_datasource; compute stats indi_pub_downloads_datasource;
create table indi_pub_downloads_year stored as parquet as create table indi_pub_downloads_year stored as parquet as
SELECT result_id, substring(us.`date`, 1,4) as `year`, sum(downloads) no_dowloads from openaire_prod_usage_stats.usage_stats us SELECT result_id, substring(us.`date`, 1,4) as `year`, sum(downloads) no_downloads from openaire_prod_usage_stats.usage_stats us
join publication on result_id=id where downloads>0 join publication on result_id=id where downloads>0
GROUP BY result_id, `year` GROUP BY result_id, `year`
order by `year` asc; order by `year` asc;
@ -480,7 +480,7 @@ order by `year` asc;
compute stats indi_pub_downloads_year; compute stats indi_pub_downloads_year;
create table indi_pub_downloads_datasource_year stored as parquet as create table indi_pub_downloads_datasource_year stored as parquet as
SELECT result_id, substring(us.`date`, 1,4) as `year`, repository_id, sum(downloads) no_dowloads from openaire_prod_usage_stats.usage_stats us SELECT result_id, substring(us.`date`, 1,4) as `year`, repository_id, sum(downloads) no_downloads from openaire_prod_usage_stats.usage_stats us
join publication on result_id=id join publication on result_id=id
where downloads>0 where downloads>0
GROUP BY result_id, repository_id, `year` GROUP BY result_id, repository_id, `year`