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

restored indi_result_org_country_collab, removed indi_result_org_collab

This commit is contained in:
Antonis Lempesis 2022-05-06 12:52:10 +03:00
parent cfbbcaf7c4
commit 61b4c19e65
1 changed files with 20 additions and 20 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

40
dhp-workflows/dhp-stats-update/src/main/resources/eu/dnetlib/dhp/oa/graph/stats/oozie_app/scripts/step16-createIndicatorsTables.sql
View File

@ -82,31 +82,31 @@ on r.id= tmp.id;
compute stats indi_funded_result_with_fundref;
-- create table indi_result_org_country_collab stored as parquet as
-- with tmp as
-- (select o.id as id, o.country , ro.id as result,r.type from organization o
-- join result_organization ro on o.id=ro.organization
-- join result r on r.id=ro.id where o.country <> 'UNKNOWN')
-- select o1.id org1,o2.country country2, o1.type, count(distinct o1.result) as collaborations
-- from tmp as o1
-- join tmp as o2 on o1.result=o2.result
-- where o1.id<>o2.id and o1.country<>o2.country
-- group by o1.id, o1.type,o2.country;
--
-- compute stats indi_result_org_country_collab;
create table indi_result_org_collab stored as parquet as
create table indi_result_org_country_collab stored as parquet as
with tmp as
(select o.id, ro.id as result,r.type from organization o
(select o.id as id, o.country , ro.id as result,r.type from organization o
join result_organization ro on o.id=ro.organization
join result r on r.id=ro.id)
select o1.id org1,o2.id org2, o1.type, count(distinct o1.result) as collaborations
join result r on r.id=ro.id where o.country <> 'UNKNOWN')
select o1.id org1,o2.country country2, o1.type, count(distinct o1.result) as collaborations
from tmp as o1
join tmp as o2 on o1.result=o2.result
where o1.id<>o2.id
group by o1.id, o2.id, o1.type;
where o1.id<>o2.id and o1.country<>o2.country
group by o1.id, o1.type,o2.country;
compute stats indi_result_org_collab;
compute stats indi_result_org_country_collab;
-- create table indi_result_org_collab stored as parquet as
-- with tmp as
-- (select o.id, ro.id as result,r.type from organization o
-- join result_organization ro on o.id=ro.organization
-- join result r on r.id=ro.id)
-- select o1.id org1,o2.id org2, o1.type, count(distinct o1.result) as collaborations
-- from tmp as o1
-- join tmp as o2 on o1.result=o2.result
-- where o1.id<>o2.id
-- group by o1.id, o2.id, o1.type;
--
-- compute stats indi_result_org_collab;
create table indi_funder_country_collab stored as parquet as
with tmp as (select funder, project, country from organization_projects op