> Hi Karl,
>> On our website we track visits and signups for each of our campaigns.
>> Visit is a unique visit to the site and a visit is associated with a
>> signup if that user signs up. I created a table for visits:
>>>> @table1 = Signup.report_table_by_sql "select count(*),c.code from
>> signups s,campaigns c where s.campaign_id=c.id and s.user_id is not
>> null group by code"
>> +---------------------+
>> | code | signup_count |
>> +---------------------+
>> | abc | 50 |
>> | dfg | 50 |
>> +---------------------+
>>>> @table2 = Visit.report_table_by_sql "select count(*),c.code from
>> visits v,campaigns c where v.campaign_id=c.id group by code"
>>>> @table2.rename_column("count(*)", 'visit_count')
>> +--------------------+
>> | code | visit_count |
>> +--------------------+
>> | abc | 300 |
>> | dfg | 300 |
>> +--------------------+
>> What I would really like is to combine these two results into one
>> table with code, signup_count, visit_count, and conversion rate.
> Have you tried using SQL sub-selects? I can never remember the syntax
> without looking it up but it's something like:
> SELECT signup.code, signup.count, visit.count FROM (SELECT count(*),
> c.code FROM signups s......) AS signup JOIN (SELECT ....) AS visit ON
> visit.code = signup.code;
> You can do this via either model, I don't think it cares what attributes
> are returned!
> Hope that helps!
> Andrew
|
