Count data as zero if it is null when where clause is used

postgresql, sql

Solution

Your where condition on the outer joined table turns the outer join into an inner join (because the "non-existing rows will have a `NULL` value and the comparison of `NULL` with something else yields "undefined" and thus will remove that row from the result)

You need to move that condition into the join condition:

SELECT opp.name as name,
       count(log.stage_id) as stage_count
FROM crm_lead as opp 
  LEFT JOIN crm_lead_stage_log as log 
         ON opp.id = log.opportunity_id
        AND log.create_date > '2014-01-28 08:49:03'
GROUP BY name;

Problem

Now I have this query: ``` SELECT opp.name as name, count(log.stage_id) as stage_count FROM crm_lead as opp LEFT OUTER JOIN crm_lead_stage_log as log ON (opp.id = log.opportunity_id) GROUP BY name ``` And it outputs this result: ``` name | stage_count | name1 | 2 name2 | 1 name3 | 0 ``` And it outputs what I need. But if I put any condition to it, then it skips rows with zero count, which I need to be able to see. For example if I write this query: ``` SELECT opp.name as name, count(log.stage_id) as stage_count FROM crm_lead as opp LEFT OUTER JOIN crm_lead_stage_log as log ON (opp.id = log.opportunity_id) WHERE WHEN log.create_date > '2014-01-28 08:49:03' GROUP BY name ``` Then it outputs this: ``` name | stage_count | name1 | 1 ``` It counts existing stages number in that time interval correctly, but it skips rows which stages number is not existing in the time inerval. How can I make it output like this (in that example only one stage for first row is counted in that time interval with new query, for other rows, it counts zero, because it does not exist): ``` name | stage_count | name1 | 1 name2 | 0 name3 | 0 ``` Is it possible to do it like that? P.S. if more information is needed, like to put this query sample online to check it out, just write a comment and I will update my answer).

Original source