everybody!
I have a table t(id, date1, date2):
id date1 date2
1 '2020-01-02' '2020-01-02'
1 '2020-01-12' '2020-01-02'
1 '2020-02-02' '2020-01-02'
1 '2020-03-02' '2020-01-02'
2 '2020-01-12' '2020-01-02'
2 '2020-01-15' '2020-01-02'
1 '2020-05-02' '2020-01-02'
1 '2020-06-02' '2020-01-02'
I need to update it like this:
id date1 date2
1 '2020-01-02' '2020-01-11'
1 '2020-01-12' '2020-02-01'
1 '2020-02-02' '2020-02-01'
1 '2020-03-02' '2020-05-01'
2 '2020-01-12' '2020-01-14'
2 '2020-01-15' '2999-12-31'
1 '2020-05-02' '2020-06-01'
1 '2020-06-02' '2999-12-31'
in rows with equal id:
date2 = date1 [from next row] - 1
and for the last date1 in group of equal id:
date2 = '2999-12-31'
assuming you can order your columns by id and date1 in order to find which is the next row, you could use the LEAD, partitioned by id to get the next item in the list.
I replicate your case with:
create table test (id int, date1 date, date2 date);
insert into test values (1,'2020-01-02','2020-01-02');
insert into test values (1,'2020-01-12','2020-01-02');
insert into test values (1,'2020-02-02','2020-01-02');
insert into test values (1,'2020-03-02','2020-01-02');
insert into test values (2,'2020-01-12','2020-01-02');
insert into test values (2,'2020-01-15','2020-01-02');
insert into test values (1,'2020-05-02','2020-01-02');
insert into test values (1,'2020-06-02','2020-01-02');
And I could fetch the following date1 for each row with
select
id,
date1,
date2,
coalesce(lead(date1, 1) OVER (PARTITION BY id ORDER BY id, date1),'2999-12-31') date2_next
from test;
Result
id | date1 | date2 | date2_next
----+------------+------------+------------
1 | 2020-01-02 | 2020-01-02 | 2020-01-12
1 | 2020-01-12 | 2020-01-02 | 2020-02-02
1 | 2020-02-02 | 2020-01-02 | 2020-03-02
1 | 2020-03-02 | 2020-01-02 | 2020-05-02
1 | 2020-05-02 | 2020-01-02 | 2020-06-02
1 | 2020-06-02 | 2020-01-02 | 2999-12-31
2 | 2020-01-12 | 2020-01-02 | 2020-01-15
2 | 2020-01-15 | 2020-01-02 | 2999-12-31
(8 rows)
If you're looking for the update statement, check the one below
update test set date2=date2_next
from
(select id,
date1,
coalesce(lead(date1, 1) OVER (PARTITION BY id ORDER BY id, date1),'2999-12-31') date2_next
from test) nxt
where test.id = nxt.id and test.date1=nxt.date1;
Related
I have two tables login_attempts and checkouts in Amazon RedShift. A user can have multiple (un)successful login attempts and multiple (un)successful checkouts as shown in this example:
login_attempts
login_id | user_id | login | success
-------------------------------------------------------
1 | 1 | 2021-07-01 14:00:00 | 0
2 | 1 | 2021-07-01 16:00:00 | 1
3 | 2 | 2021-07-02 05:01:01 | 1
4 | 1 | 2021-07-04 03:25:34 | 0
5 | 2 | 2021-07-05 11:20:50 | 0
6 | 2 | 2021-07-07 12:34:56 | 1
and
checkouts
checkout_id | checkout_time | user_id | success
------------------------------------------------------------
1 | 2021-07-01 18:00:00 | 1 | 0
2 | 2021-07-02 06:54:32 | 2 | 1
3 | 2021-07-04 13:00:01 | 1 | 1
4 | 2021-07-08 09:05:00 | 2 | 1
Given this information, how can I get the following table with historical performance included for each checkout AS OF THAT TIME?
checkout_id | checkout | user_id | lastGoodLogin | lastFailedLogin | lastGoodCheckout | lastFailedCheckout |
---------------------------------------------------------------------------------------------------------------------------------------
1 | 2021-07-01 18:00:00 | 1 | 2021-07-01 16:00:00 | 2021-07-01 14:00:00 | NULL | NULL
2 | 2021-07-02 06:54:32 | 2 | 2021-07-02 05:01:01 | NULL | NULL | NULL
3 | 2021-07-04 13:00:01 | 1 | 2021-07-01 16:00:00 | 2021-07-04 03:25:34 | NULL | 2021-07-01 18:00:00
4 | 2021-07-08 09:05:00 | 2 | 2021-07-07 12:34:56 | 2021-07-05 11:20:50 | 2021-07-02 06:54:32 | NULL
Update: I was able to get lastFailedCheckout & lastGoodCheckout because that's doing window operations on the same table (checkouts) but I am failing to understand how to best join it with login_attempts table to get last[Good|Failed]Login fields. (sqlfiddle)
P.S.: I am open to PostgreSQL suggestions as well.
Good start! A couple things in your SQL - 1) You should really try to avoid inequality joins as these can lead to data explosions and aren't needed in this case. Just put a CASE statement inside your window function to use only the type of checkout (or login) you want. 2) You can use the frame clause to not self select the same row when finding previous checkouts.
Once you have this pattern you can use it to find the other 2 columns of data you are looking for. The first step is to UNION the tables together, not JOIN. This means making a few more columns so the data can live together but that is easy. Now you have the userid and the time the "thing" happened all in the same data. You just need to WINDOW 2 more times to pull the info you want. Lastly, you need to strip out the non-checkout rows with an outer select w/ where clause.
Like this:
create table login_attempts(
loginid smallint,
userid smallint,
login timestamp,
success smallint
);
create table checkouts(
checkoutid smallint,
userid smallint,
checkout_time timestamp,
success smallint
);
insert into login_attempts values
(1, 1, '2021-07-01 14:00:00', 0),
(2, 1, '2021-07-01 16:00:00', 1),
(3, 2, '2021-07-02 05:01:01', 1),
(4, 1, '2021-07-04 03:25:34', 0),
(5, 2, '2021-07-05 11:20:50', 0),
(6, 2, '2021-07-07 12:34:56', 1)
;
insert into checkouts values
(1, 1, '2021-07-01 18:00:00', 0),
(2, 2, '2021-07-02 06:54:32', 1),
(3, 1, '2021-07-04 13:00:01', 1),
(4, 2, '2021-07-08 09:05:00', 1)
;
SQL:
select *
from (
select
c.checkoutid,
c.userid,
c.checkout_time,
max(case success when 0 then checkout_time end) over (
partition by userid
order by event_time
rows between unbounded preceding and 1 preceding
) as lastFailedCheckout,
max(case success when 1 then checkout_time end) over (
partition by userid
order by event_time
rows between unbounded preceding and 1 preceding
) as lastGoodCheckout,
max(case lsuccess when 0 then login end) over (
partition by userid
order by event_time
rows between unbounded preceding and 1 preceding
) as lastFailedLogin,
max(case lsuccess when 1 then login end) over (
partition by userid
order by event_time
rows between unbounded preceding and 1 preceding
) as lastGoodLogin
from (
select checkout_time as event_time, checkoutid, userid,
checkout_time, success,
NULL as login, NULL as lsuccess
from checkouts
UNION ALL
select login as event_time,NULL as checkoutid, userid,
NULL as checkout_time, NULL as success,
login, success as lsuccess
from login_attempts
) c
) o
where o.checkoutid is not null
order by o.checkoutid
This is TSQL and I'm trying to calculate repeat purchase rate for last 12 months. This is achieved by looking at sum of customers who have bought more than 1 time last 12 months and the total number of customers last 12 months.
The SQL code below will give me just that; but i would like to dynamically do this for the last 12 months. This is the part where i'm stuck and not should how to best achieve this.
Each month should include data going back 12 months. I.e. June should hold data between June 2018 and June 2018, May should hold data from May 2018 till May 2019.
[Order Date] is a normal datefield (yyyy-mm-dd hh:mm:ss)
DECLARE #startdate1 DATETIME
DECLARE #enddate1 DATETIME
SET #enddate1 = DATEADD(MONTH, DATEDIFF(MONTH, 0, GETDATE())-1, 0) -- Starting June 2018
SET #startdate1 = DATEADD(mm,DATEDIFF(mm,0,GETDATE())-13,0) -- Ending June 2019
;
with dataset as (
select [Phone No_] as who_identifier,
count(distinct([Order No_])) as mycount
from [MyCompany$Sales Invoice Header]
where [Order Date] between #startdate1 and #enddate1
group by [Phone No_]
),
frequentbuyers as (
select who_identifier, sum(mycount) as frequentbuyerscount
from dataset
where mycount > 1
group by who_identifier),
allpurchases as (
select who_identifier, sum(mycount) as allpurchasescount
from dataset
group by who_identifier
)
select sum(frequentbuyerscount) as frequentbuyercount, (select sum(allpurchasescount) from allpurchases) as allpurchasecount
from frequentbuyers
I'm hoping to achieve end result looking something like this:
...Dec, Jan, Feb, March, April, May, June each month holding both values for frequentbuyercount and allpurchasescount.
Here is the code. I made a little modification for the frequentbuyerscount and allpurchasescount. If you use a sumif like expression you don't need a second cte.
if object_id('tempdb.dbo.#tmpMonths') is not null drop table #tmpMonths
create table #tmpMonths ( MonthID datetime, StartDate datetime, EndDate datetime)
declare #MonthCount int = 12
declare #Month datetime = DATEADD(MONTH, DATEDIFF(MONTH, 0, GETDATE()), 0)
while #MonthCount > 0 begin
insert into #tmpMonths( MonthID, StartDate, EndDate )
select #Month, dateadd(month, -12, #Month), #Month
set #Month = dateadd(month, -1, #Month)
set #MonthCount = #MonthCount - 1
end
;with dataset as (
select m.MonthID as MonthID, [Phone No_] as who_identifier,
count(distinct([Order No_])) as mycount
from [MyCompany$Sales Invoice Header]
inner join #tmpMonths m on [Order Date] between m.StartDate and m.EndDate
group by m.MonthID, [Phone No_]
),
buyers as (
select MonthID, who_identifier
, sum(iif(mycount > 1, mycount, 0)) as frequentbuyerscount --sum only if count > 1
, sum(mycount) as allpurchasescount
from dataset
group by MonthID, who_identifier
)
select
b.MonthID
, max(tm.StartDate) StartDate, max(tm.EndDate) EndDate
, sum(b.frequentbuyerscount) as frequentbuyercount
, sum(b.allpurchasescount) as allpurchasecount
from buyers b inner join #tmpMonths tm on tm.MonthID = b.MonthID
group by b.MonthID
Be aware, that the code was tested only syntax-wise.
After the test data, this is the result:
MonthID | StartDate | EndDate | frequentbuyercount | allpurchasecount
-----------------------------------------------------------------------------
2018-08-01 | 2017-08-01 | 2018-08-01 | 340 | 3702
2018-09-01 | 2017-09-01 | 2018-09-01 | 340 | 3702
2018-10-01 | 2017-10-01 | 2018-10-01 | 340 | 3702
2018-11-01 | 2017-11-01 | 2018-11-01 | 340 | 3702
2018-12-01 | 2017-12-01 | 2018-12-01 | 340 | 3703
2019-01-01 | 2018-01-01 | 2019-01-01 | 340 | 3703
2019-02-01 | 2018-02-01 | 2019-02-01 | 2 | 8
2019-03-01 | 2018-03-01 | 2019-03-01 | 2 | 3
2019-04-01 | 2018-04-01 | 2019-04-01 | 2 | 3
2019-05-01 | 2018-05-01 | 2019-05-01 | 2 | 3
2019-06-01 | 2018-06-01 | 2019-06-01 | 2 | 3
2019-07-01 | 2018-07-01 | 2019-07-01 | 2 | 3
I'm trying to create this query to get all complete date on range and data with nulls if the date is not exist on the table
For example this is my tbl_example
Original data:
id | userid(str) | comment(str) | mydate(date)
1 0001 sample1 2019-06-20T16:00:00.000Z
2 0002 sample2 2019-06-21T16:00:00.000Z
3 0003 sample3 2019-06-24T16:00:00.000Z
4 0004 sample4 2019-06-25T16:00:00.000Z
5 0005 sample5 2019-06-26T16:00:00.000Z
Then:
select * from tbl_example where mydate between '2019-06-20' AND
DATE('2019-06-20') + interval '5 day')
how to output all the dates on range with possible null like this
Expected output:
id | userid(str) | comment(str) | mydate(date)
1 0001 sample1 2019-06-20T16:00:00.000Z
2 0002 sample2 2019-06-21T16:00:00.000Z
null null null 2019-06-22T16:00:00.000Z
null null null 2019-06-23T16:00:00.000Z
4 0003 sample3 2019-06-24T16:00:00.000Z
5 0004 sample4 2019-06-25T16:00:00.000Z
This is my sample test environment: http://www.sqlfiddle.com/#!17/f5285/2
OK, just see my SQL as below:
with all_dates as (
select generate_series(min(mydate),max(mydate),'1 day'::interval) as dates from tbl_example
)
,null_dates as (
select
a.dates
from
all_dates a
left join
tbl_example t on a.dates = t.mydate
where
t.mydate is null
)
select null as id, null as userid, null as comment, dates as mydate from null_dates
union
select * from tbl_example order by mydate;
id | userid | comment | mydate
----+--------+---------+---------------------
1 | 0001 | sample1 | 2019-06-20 16:00:00
2 | 0002 | sample1 | 2019-06-21 16:00:00
| | | 2019-06-22 16:00:00
| | | 2019-06-23 16:00:00
3 | 0003 | sample1 | 2019-06-24 16:00:00
4 | 0004 | sample1 | 2019-06-25 16:00:00
5 | 0005 | sample1 | 2019-06-26 16:00:00
(7 rows)
Or the generate_series clause you can just write the date arguments you want ,as below:
select generate_series('2019-06-20 16:00:00','2019-06-20 16:00:00'::timestamp + '5 days'::interval,'1 day'::interval) as dates
SELECT id, userid, "comment", d.mydate
FROM generate_series('2019-06-20'::date, '2019-06-25'::date, INTERVAL '1 day') d (mydate)
LEFT JOIN tbl_example ON d.mydate = tbl_example.mydate
Result
I´m having problems querying when lead() values are not within the range of current row, rows on the range's edge return null lead() values.
Let’s say I have a simple table to keep track of continuous counters
create table anytable
( wseller integer NOT NULL,
wday date NOT NULL,
wshift smallint NOT NULL,
wconter numeric(9,1) )
with the following values
wseller wday wshift wcounter
1 2016-11-30 1 100.5
1 2017-01-03 1 102.5
1 2017-01-25 2 103.2
1 2017-02-05 2 106.1
2 2015-05-05 2 81.1
2 2017-01-01 1 92.1
2 2017-01-01 2 93.1
3 2016-12-01 1 45.2
3 2017-01-05 1 50.1
and want net units for current year
wseller wday wshift units
1 2017-01-03 1 2
1 2017-01-25 2 0.7
1 2017-02-05 2 2.9
2 2017-01-01 1 11
2 2017-01-01 2 1
3 2017-01-05 1 4.9
If I use
seletc wseller, wday, wshift, wcounter-lead(wcounter) over (partition by wseller order by wseller, wday desc, wshift desc)
from anytable
where wday>='2017-01-01'
gives me nulls on the first wseller by partition. I´m using this query within a large CTE.
What am I doing wrong?
The scope of a window function takes into account conditions in the WHERE clause. Move the condition to the outer query:
select *
from (
select
wseller, wday, wshift,
wcounter- lead(wcounter) over (partition by wseller order by wday desc, wshift desc)
from anytable
) s
where wday >= '2017-01-01'
order by wseller, wday, wshift
wseller | wday | wshift | ?column?
---------+------------+--------+----------
1 | 2017-01-03 | 1 | 2.0
1 | 2017-01-25 | 2 | 0.7
1 | 2017-02-05 | 2 | 2.9
2 | 2017-01-01 | 1 | 11.0
2 | 2017-01-01 | 2 | 1.0
3 | 2017-01-05 | 1 | 4.9
(6 rows)
I'm having issues trying to wrap my head around how to extract some time series stats from my Postgres DB.
For example, I have several stores. I record how many sales each store made each day in a table that looks like:
+------------+----------+-------+
| Date | Store ID | Count |
+------------+----------+-------+
| 2017-02-01 | 1 | 10 |
| 2017-02-01 | 2 | 20 |
| 2017-02-03 | 1 | 11 |
| 2017-02-03 | 2 | 21 |
| 2017-02-04 | 3 | 30 |
+------------+----------+-------+
I'm trying to display this data on a bar/line graph with different lines per Store and the blank dates filled in with 0.
I have been successful getting it to show the sum per day (combining all the stores into one sum) using generate_series, but I can't figure out how to separate it out so each store has a value for each day... the result being something like:
["Store ID 1", 10, 0, 11, 0]
["Store ID 2", 20, 0, 21, 0]
["Store ID 3", 0, 0, 0, 30]
It is necessary to build a cross join dates X stores:
select store_id, array_agg(total order by date) as total
from (
select store_id, date, coalesce(sum(total), 0) as total
from
t
right join (
generate_series(
(select min(date) from t),
(select max(date) from t),
'1 day'
) gs (date)
cross join
(select distinct store_id from t) s
) using (date, store_id)
group by 1,2
) s
group by 1
order by 1
;
store_id | total
----------+-------------
1 | {10,0,11,0}
2 | {20,0,21,0}
3 | {0,0,0,30}
Sample data:
create table t (date date, store_id int, total int);
insert into t (date, store_id, total) values
('2017-02-01',1,10),
('2017-02-01',2,20),
('2017-02-03',1,11),
('2017-02-03',2,21),
('2017-02-04',3,30);