[
https://issues.apache.org/jira/browse/KYLIN-1693?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Shaofeng SHI updated KYLIN-1693:
--------------------------------
Fix Version/s: v1.5.3
Issue Type: New Feature (was: Wish)
I'm working on this; will be able to specify multiple group by columns in one
TopN measure, like this:
{code}
{
"name": "TOP",
"function": {
"expression": "TOP_N",
"parameter": {
"type": "column",
"value": "PRICE",
"next_parameter": {
"type": "column",
"value": "SELLER_ID",
"next_parameter": {
"type": "column",
"value": "LEAF_CATEG_ID",
"next_parameter": null
}
}
},
"configuration": {},
"returntype": "topn(100)"
}
}
{code}
The query would be :
{code}
select LEAF_CATEG_ID, seller_id, sum(price) total_price from test_kylin_fact
group by LEAF_CATEG_ID, seller_id order by total_price desc limit 200;
{code}
> Support multiple group-by columns for TOP_N meausre
> ---------------------------------------------------
>
> Key: KYLIN-1693
> URL: https://issues.apache.org/jira/browse/KYLIN-1693
> Project: Kylin
> Issue Type: New Feature
> Components: Query Engine
> Affects Versions: v1.5.1
> Reporter: JunAn Chen
> Assignee: Shaofeng SHI
> Priority: Minor
> Fix For: v1.5.3
>
>
> For this case:
> table name : "tbl"
> columns: (dim_city, dim_industry, keyword, pv)
> the "keyword" column has a large cardinality, for about ten million.
> currently I can build "top100 pv" in (dim_city), (dim_industry).
> But I also want to build "top100 pv" in (dim_city, dim_industry) and "top100
> pv of keyword" in (dim_city), (dim_industry) and (dim_city, dim_industrt).
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
