I have data organised as below in table T_FORECAST
Location Sublocation Delivery_forecast Delivery_Date Forecast_date
----------------------------------------------------------------------------------
1 1 100 2020-01-01 2019-01-01
1 2 50 2020-01-01 2019-05-01
1 1 90 2020-01-01 2019-06-01
1 2 70 2020-01-01 2019-10-01
. . .
I am trying to write a query that would output the sum of Delivery_forecast per location, Delivery_date, and Forecast_date.
in the example below, I would expect:
Location Delivery_forecast Delivery_Date Forecast_date
----------------------------------------------------------------------------------
1 100 2020-01-01 2019-01-01
1 150 2020-01-01 2019-05-01
1 140 2020-01-01 2019-06-01
1 160 2020-01-01 2019-10-01
I could find the list of lines I need using the request below but I cannot find the way to get the right sum. I believe I have to do a self join
SELECT DISTINCT f.Location, f.Delivery_Date, f.Forecast_date
FROM T_FORECAST f
Use a cumulative sum:
select f.*,
sum(delivery_forecast) over (partition by location, delivery_date order by forecast_date) as running_delivery_forecast
from T_FORECAST f;
First you want an aggregation to get the sums per location and dates (SUM(delivery_forecast) / GROUP BY LOCATION, DELIVERY_DATE, FORECAST_DATE). Then you want to show a running total (SUM OVER). Per location, I guess. The two combined:
select
location,
delivery_date,
forecast_date,
sum(delivery_forecast) as forcast_for_day,
sum(sum(delivery_forecast)) over (partition by location
order by delivery_date, forecast_date
) as forcast_cumulated
from t_forecast
group by location, delivery_date, forecast_date
order by location, delivery_date, forecast_date;
They key aspect of your question is that you want to always add values from two consecutive rows.
To do this you can use a "window frame" according you your specific ordering. See SQL Server - Over clause.
For example:
select
location,
sum(delivery_forecast) over(
partition by location, delivery_date
order by forecast_date
rows between 1 preceding and current row -- the magic is here!
) as delivery_forecast,
delivery_date,
forecast_date
from t_forecast
group by location, delivery_date, forecast_date
Related
I have table in Teradata SQL like below:
ID trans_date
------------------------
123 | 2021-01-01
887 | 2021-01-15
123 | 2021-02-10
45 | 2021-03-11
789 | 2021-10-01
45 | 2021-09-02
And I need to calculate average monthly number of transactions made by customers in a period between 2021-01-01 and 2021-09-01, so client with "ID" = 789 will not be calculated because he made transaction later.
In the first month (01) were 2 transactions
In the second month was 1 transaction
In the third month was 1 transaction
In the nineth month was 1 transactions
So the result should be (2+1+1+1) / 4 = 1.25, isn't is ?
How can I calculate it in Teradata SQL? Of course I showed you sample of my data.
SELECT ID, AVG(txns) FROM
(SELECT ID, TRUNC(trans_date,'MON') as mth, COUNT(*) as txns
FROM mytable
-- WHERE condition matches the question but likely want to
-- use end date 2021-09-30 or use mth instead of trans_date
WHERE trans_date BETWEEN date'2021-01-01' and date'2021-09-01'
GROUP BY id, mth) mth_txn
GROUP BY id;
Your logic translated to SQL:
--(2+1+1+1) / 4
SELECT id, COUNT(*) / COUNT(DISTINCT TRUNC(trans_date,'MON')) AS avg_tx
FROM mytable
WHERE trans_date BETWEEN date'2021-01-01' and date'2021-09-01'
GROUP BY id;
You should compare to Fred's answer to see which is more efficent on your data.
I have a dataset in bigquery which contains order_date: DATE and customer_id.
order_date | CustomerID
2019-01-01 | 111
2019-02-01 | 112
2020-01-01 | 111
2020-02-01 | 113
2021-01-01 | 115
2021-02-01 | 119
I try to count distinct customer_id between the months of the previous year and the same months of the current year. For example, from 2019-01-01 to 2020-01-01, then from 2019-02-01 to 2020-02-01, and then who not bought in the same period of next year 2020-01-01 to 2021-01-01, then 2020-02-01 to 2021-02-01.
The output I am expect
order_date| count distinct CustomerID|who not buy in the next period
2020-01-01| 5191 |250
2020-02-01| 4859 |500
2020-03-01| 3567 |349
..........| .... |......
and the next periods shouldn't include the previous.
I tried the code below but it works in another way
with customers as (
select distinct date_trunc(date(order_date),month) as dates,
CUSTOMER_WID
from t
where date(order_date) between '2018-01-01' and current_date()-1
)
select
dates,
customers_previous,
customers_next_period
from
(
select dates,
count(CUSTOMER_WID) as customers_previous,
count(case when customer_wid_next is null then 1 end) as customers_next_period,
from (
select prev.dates,
prev.CUSTOMER_WID,
next.dates as next_dates,
next.CUSTOMER_WID as customer_wid_next
from customers as prev
left join customers
as next on next.dates=date_add(prev.dates,interval 1 year)
and prev.CUSTOMER_WID=next.CUSTOMER_WID
) as t2
group by dates
)
order by 1,2
Thanks in advance.
If I understand correctly, you are trying to count values on a window of time, and for that I recommend using window functions - docs here and here a great article explaining how it works.
That said, my recommendation would be:
SELECT DISTINCT
periods,
COUNT(DISTINCT CustomerID) OVER 12mos AS count_customers_last_12_mos
FROM (
SELECT
order_date,
FORMAT_DATE('%Y%m', order_date) AS periods,
customer_id
FROM dataset
)
WINDOW 12mos AS ( # window of last 12 months without current month
PARTITION BY periods ORDER BY periods DESC
ROWS BETWEEN 12 PRECEEDING AND 1 PRECEEDING
)
I believe from this you can build some customizations to improve the aggregations you want.
You can generate the periods using unnest(generate_date_array()). Then use joins to bring in the customers from the previous 12 months and the next 12 months. Finally, aggregate and count the customers:
select period,
count(distinct c_prev.customer_wid),
count(distinct c_next.customer_wid)
from unnest(generate_date_array(date '2020-01-01', date '2021-01-01', interval '1 month')) period join
customers c_prev
on c_prev.order_date <= period and
c_prev.order_date > date_add(period, interval -12 month) left join
customers c_next
on c_next.customer_wid = c_prev.customer_wid and
c_next.order_date > period and
c_next.order_date <= date_add(period, interval 12 month)
group by period;
I have a table like this, with column names as Date of Sale and insurance Salesman Names -
Date of Sale | Salesman Name | Sale Amount
2021-03-01 | Jack | 40
2021-03-02 | Mark | 60
2021-03-03 | Sam | 30
2021-03-03 | Mark | 70
2021-03-02 | Sam | 100
I want to do a group by, using the date of sale. The next column should display the cumulative count of the sellers who have made the sale till that date. But same sellers shouldn't be considered again.
For example,
The following table is incorrect,
Date of Sale | Count(Salesman Name) | Sum(Sale Amount)
2021-03-01 | 1 | 40
2021-03-02 | 3 | 200
2021-03-03 | 5 | 300
The following table is correct,
Date of Sale | Count(Salesman Name) | Sum(Sale Amount)
2021-03-01 | 1 | 40
2021-03-02 | 3 | 200
2021-03-03 | 3 | 300
I am not sure how to frame the SQL query, because there are two conditions involved here, cumulative count while ignoring the duplicates. I think the OVER clause along with the unbounded row preceding may be of some use here? Request your help
Edit - I have added the Sale Amount as a column. I need the cumulative sum for the Sales Amount also. But in this case , all the sale amounts should be considered unlike the salesman name case where only unique names were being considered.
One approach uses a self join and aggregation:
WITH cte AS (
SELECT t1.SaleDate,
COUNT(CASE WHEN t2.Salesman IS NULL THEN 1 END) AS cnt,
SUM(t1.SaleAmount) AS amt
FROM yourTable t1
LEFT JOIN yourTable t2
ON t2.Salesman = t1.Saleman AND
t2.SaleDate < t1.SaleDate
GROUP BY t1.SaleDate
)
SELECT
SaleDate,
SUM(cnt) OVER (ORDER BY SaleDate) AS NumSalesman,
SUM(amt) OVER (ORDER BY SaleDate) AS TotalAmount
FROM cte
ORDER BY SaleDate;
The logic in the CTE is that we try to find, for each salesman, an earlier record for the same salesman. If we can't find such a record, then we assume the record in question is the first appearance. Then we aggregate by date to get the counts per day, and finally take a rolling sum of counts in the outer query.
The best way to do this uses window functions to determine the first time a sales person appears. Then, you just want cumulative sums:
select saledate,
sum(case when seqnum = 1 then 1 else 0 end) over (order by saledate) as num_salespersons,
sum(sum(sales)) over (order by saledate) as running_sales
from (select t.*,
row_number() over (partition by salesperson order by saledate) as seqnum
from t
) t
group by saledate
order by saledate;
Note that this in addition to being more concise, this should have much, much better performance than a solution that uses a self-join.
I have a table called table1:
id created_date
1001 2020-06-01
1001 2020-01-01
1001 2020-07-01
1002 2020-02-01
1002 2020-04-01
1003 2020-09-01
I'm trying to write a query that provides me a list of distinct IDs with the earliest created_date they have, along with the count of rows each id has:
id created_date count
1001 2020-01-01 3
1002 2020-02-01 2
1003 2020-09-01 1
I managed to write a window function to grab the earliest date, but I'm having trouble figuring out where to fit the count statement in one:
SELECT
id,
created_date
FROM ( SELECT
id,
created_date,
row_number() OVER(PARTITION BY id ORDER BY created_date) as row_num
FROM table1)
) AS a
WHERE row_num = 1
You would use aggregation:
select id, min(create_date), count(*)
from table1
group by id;
I find it amusing that you want to use window functions -- which are considered more advanced -- when lowly aggregation suffices.
Assuming this data:
ID DATE
44 2019-12-31
45 2020-01-01
46 2020-01-01
47 2020-01-02
48 2020-01-03
48 2020-01-03
48 2020-01-03
How do I make a query that returns something like this?
TOTAL DATE
2 2020-01-01
3 2020-01-02
6 2020-01-03
I want all entries after a certain data, but with a counter that starts with the number of entries on the first day, then for every day, it adds the number of entries on that day. I want to plot them on a chart that shows the speed of the growth.
Is this possible? I'm on PostgreSQL 10.
Thanks!
You can use aggregation and window functions:
select date,
count(*) as count_on_date,
sum(count(*)) over (order by min(date)) as running_count
from t
where date >= '2020-01-01'
group by date
order by date;
If you wanted a count going back in time, then you would use:
select greatest(date, '2020-01-01'::date) as date
count(*) as count_on_date,
sum(count(*)) over (order by min(date)) as running_count
from t
group by greatest(date, '2020-01-01'::date)
order by greatest(date, '2020-01-01'::date);