numbers of users buying the exact same product from the same shop for > 2 times in 1 years - sql

Question

I have data like this:
date user prod shop cat1 cat2
2022-02-01 1 a a ah g
2022-02-02 1 a1 b ah g
2022-04-03 1 a a ah g
2022-04-19 1 a a ah g
2022-05-01 2 b c bg g
I want to know how many user buy the same product in the same shop for >2 times in period 1 year. The result i want like:
table 1
cat1 number_of_user
ah 1
table 2
cat2 number_of_user
g 1
For total user, my query like:
WITH data_product AS(
SELECT DATE(payment_time) date,
user,
CONCAT(prod, "_", shop) product_shop,
cat1,
cat2
FROM
a
WHERE
DATE(payment_time) BETWEEN "2022-01-01" AND DATE_SUB(current_date, INTERVAL 1 day)
ORDER BY 1,2,3),
purchased AS (
SELECT user, product_shop, count(product_shop) tot_purchased
FROM data_product
GROUP BY 1,2
HAVING COUNT(product_shop) > 2
)
SELECT COUNT(user) number_of_user FROM purchased
Please help to get number of user buy the same product in the same shop more than 2 times in period based on cat1 and cat2.

Answer 1

Try this:
create temporary table table1 as(
select *,extract(YEAR from date) as year from `projectid.dataset.table`
);
create temporary table table2 as(
select * except(date,cat2) ,count(user) over(partition by cat1,year,user,prod,shop) tcount from table1
);
create temporary table table4 as(
select * except(date,cat1) ,count(user) over(partition by cat2,year,user,prod,shop) tcount from table1
);
select distinct year,cat1 ,count(distinct user) number_of_user from table2 where tcount>2 group by YEAR,cat1;
select distinct year,cat2 ,count(distinct user) number_of_user from table4 where tcount>2 group by YEAR,cat2;
If you want a single result set you can union both the select statements.

Answer 2

I think this query might work. The first part shows count of customers who purchased same product in category1 from same shop during one year. Second part shows that for category2, then we concatenate the two set by union operation :
with cte as
(select distinct
PDate,userID as userID,prod as prod,shop,cat1 as cat1,cat2,
count(userID) over (partition by UserID,prod,shop,year(Pdate),cat1) as cat1_count,
count(PDate) over (partition by UserID,prod,shop,year(Pdate),cat2) as cat2_count
from tbl1)
select
cte.cat1 as c1,'0' as c2,count(distinct cte.cat1) as Num
from cte
where cte.cat1_count>1
group by cte.prod,cte.userID,cte.cat1
union
select
'0',cte.cat2,count(distinct cte.cat2)
from cte
where cte.cat2_count>1
group by cte.prod,cte.userID,cte.cat2