abundant-air-78515
01/30/2024, 9:37 PMfresh-football-47124
fresh-football-47124
abundant-air-78515
01/31/2024, 6:40 AMabundant-air-78515
01/31/2024, 6:41 AMabundant-air-78515
01/31/2024, 6:42 AMSELECT
21 AS value,
3c.created_at AS timestamp,
4c.employee_account_id AS user_id,
5c.company_id as company_id
6FROM
7 demando-data.prod_public.connection c
8 INNER JOIN demando-data.prod_public.account AS a ON a.id = c.employee_account_id
9 WHERE
10 c.assistant_id IS NULL
11AND c.initiated_by = 'employee'
abundant-air-78515
02/01/2024, 7:37 PMfresh-football-47124
abundant-air-78515
02/01/2024, 7:40 PMabundant-air-78515
02/01/2024, 7:40 PMabundant-air-78515
02/01/2024, 7:40 PMabundant-air-78515
02/02/2024, 12:32 PMabundant-air-78515
02/02/2024, 12:43 PMabundant-air-78515
02/15/2024, 2:25 PMhelpful-application-7107
02/15/2024, 10:13 PMCOALESCE(SUM(value, 0))
, so if you have NULLs in your data and you have SUM(value)
as your aggregation then you can get different results.abundant-air-78515
02/18/2024, 1:23 PMabundant-air-78515
02/18/2024, 1:24 PMabundant-air-78515
02/18/2024, 1:25 PMabundant-air-78515
02/18/2024, 1:26 PMabundant-air-78515
02/18/2024, 1:29 PMhelpful-application-7107
02/20/2024, 5:52 PM1
, not SUM(COALESCE(value, 0))
abundant-air-78515
02/20/2024, 5:56 PMhelpful-application-7107
02/20/2024, 5:57 PMabundant-air-78515
02/21/2024, 6:57 AMabundant-air-78515
02/21/2024, 6:57 AMhelpful-application-7107
02/21/2024, 4:32 PMUser Value Aggregation
to the number 1
. This is what I have been asking you to do.abundant-air-78515
02/22/2024, 6:38 AMabundant-air-78515
02/22/2024, 6:38 AMabundant-air-78515
02/22/2024, 6:38 AMhelpful-application-7107
02/22/2024, 5:56 PMhelpful-application-7107
02/22/2024, 5:56 PMabundant-air-78515
02/22/2024, 6:41 PMhelpful-application-7107
02/22/2024, 6:42 PMhelpful-application-7107
02/22/2024, 6:42 PMabundant-air-78515
02/22/2024, 6:42 PMabundant-air-78515
02/22/2024, 6:43 PMabundant-air-78515
02/22/2024, 6:44 PMabundant-air-78515
02/22/2024, 6:44 PMabundant-air-78515
02/22/2024, 6:44 PMabundant-air-78515
02/22/2024, 6:44 PMhelpful-application-7107
02/22/2024, 6:45 PMnow i just want to count accept in one exeperientWhat do you mean by this?
abundant-air-78515
02/22/2024, 6:46 PMabundant-air-78515
02/22/2024, 6:46 PMabundant-air-78515
02/22/2024, 6:47 PMabundant-air-78515
02/22/2024, 6:47 PMhelpful-application-7107
02/22/2024, 6:47 PMabundant-air-78515
02/22/2024, 6:51 PMhelpful-application-7107
02/22/2024, 6:53 PM1 as value
in the query.
User Value Aggregation: leave it blank so it accepts the default, you want basically to sum the rows and the default will do that if you have 1 as value
Denominator: do not pick one. We will use all experiment users as default
Metric 2 (count accepted):
Type: count
SQL: one row per accepted request, with 1 as value
in the query.
User Value Aggregation: again leave it blank
Denominator: do not pick one. We will use all experiment users as default
Metric 3 (ratio accepted / sent):
Type: count (not binomial, that's something different where each user is either a 0/1)
SQL: same as count accepted above
Denominator: count sent
should be picked as your denominatorhelpful-application-7107
02/22/2024, 6:54 PMabundant-air-78515
02/22/2024, 6:56 PMabundant-air-78515
02/22/2024, 6:56 PMabundant-air-78515
02/22/2024, 6:56 PMabundant-air-78515
02/22/2024, 6:57 PMabundant-air-78515
02/22/2024, 6:57 PMhelpful-application-7107
02/22/2024, 6:58 PMsent / accepted
which would be larger than 1.
But at the end of the day it's because binomial forces users to be either 0 or 1, which isn't what you want here.helpful-application-7107
02/22/2024, 6:58 PMhelpful-application-7107
02/22/2024, 6:58 PM% of users to accept at least one request
then you could use binomial.abundant-air-78515
02/22/2024, 6:58 PMabundant-air-78515
02/22/2024, 6:58 PMhelpful-application-7107
02/22/2024, 6:58 PMhelpful-application-7107
02/22/2024, 6:59 PMabundant-air-78515
02/22/2024, 6:59 PM