I'm having a bit of issue trying to sum two column together with a left join - running in to grouping problems.
An example of the issue is below:
Table One: [Order]
ID CustomerID
1 512
2 317
3 562
Table Two: [OrderEntry]
OrderID Type ID QuantitySold QuantityReturned
1 A 1 1 0
1 A 2 3 0
1 A 3 1 1
2 A 4 1 1
3 B 5 2 0
What I'm trying to display:
CustomerID ID Sold - Returned
512 1 1
512 1 3
512 1 0
317 2 0
Where [OrderEntry].Type = 'A'
This is very basic SQL:
SELECT
ord.CustomerID
, ord.ID
, orden.QuantitySold - orden.QuantityReturned AS [Sold - Returned]
FROM Order ord
LEFT JOIN OrderEntry orden
ON ord.ID = orden.ID
WHERE orden.Type = 'A'
Here you can use any join as you are using and use concat function on two of your column like this
select concat(OrderEntry.QuantitySold, OrderEntry.QuantityReturned) AS newcolumn_name
Related
I seek to find the maximum timestamp (ob.create_ts) for each group of marketid's (ob.marketid), joining tables obe (ob.orderbookid = obe.orderbookid) and market (ob.marketid = m.marketid). Although there are a number of solutions posted like this for a single table, when I join multiple tables, I get redundant results. Sample table and desired results below:
table: ob
orderbookid
marketid
create_ts
1
1
1664635255298
2
1
1664635255299
3
1
1664635255300
4
2
1664635255301
5
2
1664635255302
6
2
1664635255303
table: obe
orderbookentryid
orderbookid
entryname
1
1
'entry-1'
2
1
'entry-2'
3
1
'entry-3'
4
2
'entry-4'
5
2
'entry-5'
6
3
'entry-6'
7
3
'entry-7'
8
4
'entry-8'
9
5
'entry-9'
10
6
'entry-10'
table: m
marketid
marketname
1
'market-1'
2
'market-2'
desired results
ob.orderbookid
ob.marketid
obe.orderbookentryid
obe.entryname
m.marketname
3
1
6
'entry-6'
'market-1'
3
1
7
'entry-7'
'market-1'
6
2
10
'entry-10'
'market-2'
Use ROW_NUMBER() to get a properly filtered ob table. Then JOIN the other tables onto that!
WITH
ob_filtered AS (
SELECT
orderbookid,
marketid
FROM
(
SELECT
*,
ROW_NUMBER() OVER (
PARTITION BY
marketid
ORDER BY
create_ts DESC
) AS create_ts_rownumber
FROM
ob
) ob_with_rownumber
WHERE
create_ts_rownumber = 1
)
SELECT
ob_filtered.orderbookid,
ob_filtered.marketid,
obe.orderbookentryid,
obe.entryname,
m.marketname
FROM
ob_filtered
JOIN m
ON m.marketid = ob_filtered.marketid
JOIN obe
ON ob_filtered.orderbookid = obe.orderbookid
;
I'd like to count total events, which can have two different values, and I could not figure out how to merge them together. My query is the following:
SELECT TOP(20)
[MatchEvents].[PlayerID], [MatchEvents].[EventType],
COUNT([MatchEvents].[ID]) AS [TOTAL]
FROM
[MatchEvents]
INNER JOIN
[Match] ON [MatchEvents].[MatchID] = [Match].[ID]
AND [Match].[Season] = 1
WHERE
([MatchEvents].[EventType] = 0 OR [MatchEvents].[EventType] = 1)
GROUP BY
[MatchEvents].[PlayerID], [MatchEvents].[EventType]
ORDER BY
[TOTAL] ESC
Current output:
PlayerID
EventType
Total
1
0
8
1
1
3
2
0
8
2
1
3
3
0
8
3
1
3
Expected output:
PlayerID
Total
1
11
2
11
3
11
How could I merge my current results further?
Thanks!
From your expected results it appears you just need to remove grouping by EventType
I would suggest the following:
select top(20) me.PlayerID, Count(*) as Total
from MatchEvents me
join [Match] m on m.Id = me.MatchId and m.Season = 1
where me.EventType in (0, 1)
group by me.PlayerID
order by Total desc;
I just started learning SAS and realised that proc sql don't use window functions. As I am more at ease with sql I was wondering how I can simulate a sum window function in proc?
desired result
select a.active, a.store_id, a.nbr, sum(nbr) over (partition by a.store_id)
from(select active, store_id, count(customer_id) as nbr from customer group by active, store_id) as a
;
active
store_id
nbr
sum
0
1
8
326
1
1
318
326
0
2
7
273
1
2
266
273
eg of raw data
select active, store_id, customer_id
from customer
limit 10;
active
store_id
customer_id
1
1
1
1
1
2
1
2
3
1
2
4
1
1
5
1
1
6
0
1
7
1
2
8
1
1
9
1
2
10
current result and query
select a.active, a.store_id, a.nbr, sum(nbr)
from(select active, store_id, count(customer_id) as nbr from customer group by active, store_id) as a
group by a.active, a.store_id, a.nbr;
active
store_id
nbr
sum
0
1
8
8
1
1
318
318
0
2
7
7
1
2
266
266
Unlike some SQL implementation SAS is happy to re-merge summary statistics back onto the detail rows when you include variables that are neither group by nor summary statistics.
Let's convert your print outs of data into an actual dataset. And let's change one value so we have at least two values of ACTIVE to group by.
data have;
input active store_id customer_id;
cards;
1 1 1
1 1 2
1 2 3
1 2 4
1 1 5
1 1 6
0 1 7
1 2 8
1 1 9
1 2 10
;
Now we can count the records by ACTIVE and STORE_ID and then generate the report by appending the store total.
proc sql;
select active,store_id,nbr,sum(nbr) as store_nbr
from (select active,store_id,count(*) as nbr
from have
group by active,store_id
)
group by store_id
;
Resulting printout:
active store_id nbr store_nbr
---------------------------------------
0 1 1 6
1 1 5 6
1 2 4 4
You can do the equivalent in proc sql by merging two sub-queries: one for the count of customers by active, store_id, and another for the total customers for each store_id.
proc sql noprint;
create table want as
select t1.active
, t1.store_id
, t1.nbr
, t2.sum
from (select active
, store_id
, count(customer_id) as nbr
from have
group by store_id, active
) as t1
LEFT JOIN
(select store_id
, count(customer_id) as sum
from have
group by store_id
) as t2
ON t1.store_id = t2.store_id
;
quit;
If you wanted to do this in a more SASsy way, you can run proc means and merge together the results from a single dataset that holds everything you need. proc means will calculate all possible combinations of your variables by default.
proc means data=have noprint;
class store_id active;
ways 1 2;
output out=want_total
n(customer_id) = total
;
run;
data want;
merge want_total(where=(_TYPE_ = 3) rename=(total = nbr) )
want_total(where=(_TYPE_ = 2) rename=(total = sum) keep=_TYPE_ store_id total)
;
by store_id;
drop _:;
run;
Or, in SQL:
proc sql;
create table want as
select t1.store_id
, t1.active
, t1.total as nbr
, t2.total as sum
from want_total as t1
LEFT JOIN
want_total as t2
ON t1.store_id = t2.store_id
where t1._TYPE_ = 3
AND t2._TYPE_ = 2
;
quit;
The _TYPE_ variable identifies the level of the analysis. For example, _TYPE_ = 1 is for active only, _TYPE_ = 2 is for store_id only, and _TYPE_ = 3 is for all combinations. You can view this in the output dataset from proc means:
store_id active _TYPE_ _FREQ_ total
. 0 1 3 3
. 1 1 7 7
1 . 2 6 6
2 . 2 4 4
1 0 3 1 1
1 1 3 5 5
2 0 3 2 2
2 1 3 2 2
And if you wanted faster high-performance results, check out its big sibling, proc hpsummary.
Therein lies the cool thing about SAS: You can bounce between PROCs, SQL, the DATA Step, and Python via Pandas/proc python. You can exploit the unique benefits of each of these methods and thought processes for any number of data engineering and statistics problems.
I have 3 tables which maintains stock entries for each products/items. These three tables like below :
Table : ItemStock (to maintain remaining stock of each item)
Id ItemId OpgQty BranchID CurrentStock
1 7 0 1 8
2 7 0 2 3
3 6 0 1 2
4 6 0 2 0
Table : ItemPurchase (StockIn)
Id ItemId Qty BranchID
1 7 5 1
2 7 4 2
3 7 6 1
4 7 2 2
5 6 4 1
6 6 2 2
7 6 2 1
Table : ItemSale (StockOut)
Id ItemId Qty BranchID
1 7 2 1
2 7 3 2
3 7 1 1
4 6 4 1
5 6 2 2
Desired Output (based on sql queries)
I want to have result like below : (part of report)
Id ItemId OpgQty BranchId StockIn StockOut CurrentStock
1 7 0 1 11 3 8
2 7 0 2 6 3 3
3 6 0 1 6 4 2
4 6 0 2 2 2 0
I was trying to get the desired result but was not able to do so. Please help!!!
try this;
select
m.Id,
m.ItemId,
m.OpgQty,
m.BranchID,
si.StockIn,
m.CurrentStock-si.StockIn StockOut,
m.CurrentStock
from
ItemStock m
inner join
(
select
ItemId,BranchId,sum(Qty) as StockIn
from
ItemPurchase
group by ItemId,BranchId
) si on si.ItemId=m.ItemId and si.BranchId=m.BranchId
A very simple query that gives the desired result is :
select *,
(select sum(Qty)
from ItemPurchase
where ItemPurchase.ItemId = ItemStock.ItemId and
ItemPurchase.BranchId = ItemStock.BranchId) as StockIn,
(select sum(Qty)
from ItemSale
where ItemSale.ItemId = ItemStock.ItemId and
ItemSale.BranchId = ItemStock.BranchId) as StockOut
from ItemStock
Two subqueries with group by and aggregation will get what you want.
select
s.*,
coalesce([ip].StockIn, 0) as StockIn, -- In case of no records in ItemPurchase or ItemSale, coalesce is neccessary.
coalesce([is].StockOut, 0) as StockOut
from ItemStock s
left join (
select sum(Qty) as StockIn, ItemId, BranchId
from ItemPurchase
group by ItemId, BranchId
) [ip] on s.ItemId = [ip].ItemId and s.BranchId = [ip].BranchId
left join (
select sum(Qty) as StockOut, ItemId, BranchId
from ItemSale
group by ItemId, BranchId
) [is] on s.ItemId = [is].ItemId and s.BranchId = [is].BranchId
See demo in sqlfiddle.
Please
Try This ... I hope you consider this too.
Have the following table tblTrans where
Trans_ID Trans Sequence Trans_PointsEarned Trans_PointsApplied
4452 1 1 1
4452 2 1 1
4452 3 0 1
4462 1 1 1
4462 2 1 1
4462 3 1 1
4462 4 1 1
4462 5 1 1
9101 1 0 1
9101 2 0 1
9101 3 0 1
9101 4 0 1
(useless table doesnt work)
I need to set the following on another field per every customer ID.
So Customer_OverallPoints
4452 = 2 (doesn't count 0's)
4462 = 4 (I want to cap the points to 4 based on the sequence and transID and customerID)
9101 = 0 (dont count 0's).
This needs to be applied to thousands of records based on customerID and TransID where Trans_Sequence is within the same Trans_ID and it only counts the first 4 rows that have the Trans_pointsEarned = 1.
I tried putting a psuedocode together but it just looked ridicilous and I can't even come up with the logic for this.
Thanks
Assuming that TransId is really the customer id, I think the basic logic is just an aggregation:
select t.TransId,
(case when sum(t.Trans_PointsEarned) > 4 then 4
else sum(t.Trans_PointsEarned)
end) as Customer_OverallPoints
from tblTrans t
group by t.TransId;
You can put this into an update statement as:
update customers c
set Customer_OverallPoints = (select (case when sum(t.Trans_PointsEarned) > 4 then 4
else sum(t.Trans_PointsEarned)
end)
from tblTrans t
where t.TransId = c.CustomerId
);