SQL: Creating a column depending on whether there is at least one value in another column that fulfi-CodePudding

I want to know whether a marketing campaign is active or not, and the only way to know it is checking if at least one of the ads is still active. So I have this table:

 ------------- -------- -------------- 
| campaign_id | ad_id  |   ad_status  |
 ------------- -------- -------------- 
|  campaign1  |  ad1   |   INACTIVE   |
 ------------- -------- -------------- 
|  campaign1  |  ad2   |   ACTIVE     |
 ------------- -------- -------------- 
|  campaign2  |  ad3   |   INACTIVE   |
 ------------- -------- -------------- 
|  campaign2  |  ad4   |   INACTIVE   |
 ------------- -------- --------------

The desired outcome would be like this:

 ------------- -------- -------------- ----------------- 
| campaign_id | ad_id  |   ad_status  | campaign_status |
 ------------- -------- -------------- ----------------- 
|  campaign1  |  ad1   |   INACTIVE   |      ACTIVE     |
 ------------- -------- -------------- ----------------- 
|  campaign1  |  ad2   |   ACTIVE     |      ACTIVE     |      
 ------------- -------- -------------- ----------------- 
|  campaign2  |  ad3   |   INACTIVE   |     INACTIVE    |
 ------------- -------- -------------- ----------------- 
|  campaign2  |  ad4   |   INACTIVE   |     INACTIVE    |
 ------------- -------- -------------- -----------------

I've been trying combining CASE WHEN and WHERE ANY inside another SELECT clause but I don't seem to get how to get it done, also I don't want to have it GROUP BY, just want that column with the repeated values.

Also I'm using BigQuery in case there is a specific BigQuery way to do this.

CodePudding user response：

Try this:

with cte as
  (
              select 'campaign1' as campaign_id, 'ad1' as ad_id, 'INACTIVE' as ad_status
    union all select 'campaign1' as campaign_id, 'ad2' as ad_id, 'ACTIVE' as ad_status
    union all select 'campaign2' as campaign_id, 'ad3' as ad_id, 'INACTIVE' as ad_status
    union all select 'campaign2' as campaign_id, 'ad4' as ad_id, 'INACTIVE' as ad_status
  )
select  t.*,
        case when (count(case when t.ad_status = "ACTIVE" then 1 else null end) over(partition by t.campaign_id)) > 0 then 'ACTIVE' else 'INACTIVE' end as campaign_status
from    cte t

CodePudding user response：

Try this by using UNNEST():

with sample_data as (
  select 'campaign1' as campaign_id, 'ad1' as ad_id, 'INACTIVE' as ad_status,
  union all select 'campaign1' as campaign_id, 'ad2' as ad_id, 'ACTIVE' as ad_status,
  union all select 'campaign2' as campaign_id, 'ad3' as ad_id, 'INACTIVE' as ad_status,
  union all select 'campaign2' as campaign_id, 'ad4' as ad_id, 'INACTIVE' as ad_status,
)

  SELECT *, if ('ACTIVE' IN UNNEST(array_agg(ad_status) over (partition by campaign_id)), 'ACTIVE', 'INACTIVE' ) AS contains_value

from sample_data

Here is the ouptut: