I am trying to create database query, which will select rows from one table, create periods from those rows (using Lag window function), and join the query with rows from different table, where it sums value(s)from one column per row in first table.
Table A:
id,
created_at,
object_id
Table B:
id,
end_time,
value,
object_id
And rows, that query yields should consist of columns something like:
lag(tablea.created_at) over(tablea.object_id, tablea.created_at),
tablea.created_at,
tablea.object_id
sum(tableb.value) where it sums the tableb.value from matching period
I tried creating query where i put the window function into WHERE clause only to get an error. I also tried puting the period into join on clause but that also raised an error.
It is no problem, if it is not possible. I just want to know if it possible and in that case how it is possible. If it is not possible, then i just will try to come up with alternative aproach.
Edit: link to example sqlfiddle: http://sqlfiddle.com/#!1/c7878
Edit2: The SQL i tried was something like:
SELECT lag(a.created_at, a.created_at, a.object_id, sum(b.value) from tablea a left join tableb b on (something) order by a.object_id, a.created_at
But obviously that did not work, because i could not use window function in ON clause. That's where i got stuck
Related
I have two queries that I'm running separately from a PHP script. The first is checking if an identifier (group) has a timestamp in a table.
SELECT
group, MAX(timestamp) AS timestamp, value
FROM table_schema.sample_table
GROUP BY group, value
If there is no timestamp, then it runs this second query that retrieves the minimum timestamp from a separate table:
SELECT
group, MIN(timestamp) as timestamp, value AS value
FROM table_schema.src_table
GROUP BY group, value
And goes on from there.
What I would like to do, for the sake of conciseness, is to have a single query that runs the first statement, but that defaults to the second if NULL. I've tried with coalesce() and CASE statements, but they require subqueries to return single columns (which I hadn't run into being an issue yet). I then decided I should try a JOIN on the table with the aggregate timestamp to get the whole row, but then quickly realized I can't variate the table being joined (not to my knowledge). I opted to try joining both results and getting the max, something like this:
Edit: I am so tired, this should be a UNION, not a JOIN
sorry for any possible confusion :(
SELECT smpl.group, smpl.value, MAX(smpl.timestamp) AS timestamp
FROM table_schema.sample_table as smpl
INNER JOIN
(SELECT src.group, src.value, MIN(src.timestamp) AS timestamp
FROM source_table src
GROUP BY src.group, src.value) AS history
ON
smpl.group = history.group
GROUP BY smpl.group, smpl.value
I don't have a SELECT MAX() on this because it's really slow as is, most likely because my SQL is a bit rusty.
If anyone knows a better approach, I'd appreciate it!
Please try this:
select mx.group,(case when mx.timestamp is null then mn.timestamp else mx.timestamp end)timestamp,
(case when mx.timestamp is null then mn.value else mx.value end)value
(
SELECT
group, MAX(timestamp) AS timestamp, value
FROM table_schema.sample_table
GROUP BY group, value
)mx
left join
(
SELECT
group, MIN(timestamp) as timestamp, value AS value
FROM table_schema.src_table
GROUP BY group, value
)mn
on mx.group = mn.group
I have a table BAR_DATA with two fields: LongDate, Time. Both are long integers. No Access Date/Time involved here.
For each distinct LongDate value there are hundreds of records, each with Time value which may be distinct or duplicate within that LongDate.
I need to create an SQL statement that will group by LongDate and give me a count of distinct Times within each LongDate.
The following SQL statement, (built by an Acess query) does NOT work (some LongDates are omitted):
Query A
SELECT DISTINCT BAR_DATA.LongDate, Count(BAR_DATA.Time) AS CountOfTime
FROM BAR_DATA
GROUP BY BAR_DATA.LongDate
HAVING (((Count(BAR_DATA.Time))<>390 And (Count(BAR_DATA.Time))<>210));
However, if I use Query B to reference Query DistinctDateTime, it does work:
Query B
SELECT DistinctDateTime.LongDate, Count(DistinctDateTime.Time) AS CountOfTime
FROM DistinctDateTime
GROUP BY DistinctDateTime.LongDate
HAVING (((Count(DistinctDateTime.Time))<>390 And (Count(DistinctDateTime.Time))<>210));
Query DistinctDateTime
SELECT DISTINCT BAR_DATA.LongDate, BAR_DATA.Time
FROM BAR_DATA;
My problem:
I need to get Query B and Query DistinctDateTime wrapped into a single SQL statement so I can paste it into a VBA function. I presume there
is some subquery techniques, but I have failed at every attempt, and find no pertinent example.
Any help will be greatly appreciated. Thanks!
Subquery your distinct table inside and perform your aggregates outside until you get the desired result:
SELECT DistinctDateTime.LongDate, Count(DistinctDateTime.Time) AS CountOfTime
FROM
(
SELECT DISTINCT BAR_DATA.LongDate, BAR_DATA.Time
FROM BAR_DATA
) AS DistinctDateTime
GROUP BY DistinctDateTime.LongDate
HAVING (((Count(DistinctDateTime.Time))<>390 And (Count(DistinctDateTime.Time))<>210));
First time posting here, hopes it goes well.
I try to make a query with Oracle SQL Developer, where it returns a customer_ID from a table and the time of the payment from another. I'm pretty sure that the problems lies within my logicflow (It was a long time I used SQL, and it was back in school so I'm a bit rusty in it). I wanted to list the IDs as DISTINCT and ORDER BY the dates ASCENDING, so only the first date would show up.
However the returned table contains the same ID's twice or even more in some cases. I even found the same ID and same DATE a few times while I was scrolling through it.
If you would like to know more please ask!
SELECT DISTINCT
FIRM.customer.CUSTOMER_ID,
FIRM.account_recharge.X__INSDATE FELTOLTES
FROM
FIRM.customer
INNER JOIN FIRM.account
ON FIRM.customer.CUSTOMER_ID = FIRM.account.CUSTOMER
INNER JOIN FIRM.account_recharge
ON FIRM.account.ACCOUNT_ID = FIRM.account_recharge.ACCOUNT
WHERE
FIRM.account_recharge.X__INSDATE BETWEEN TO_DATE('14-01-01', 'YY-MM-DD') AND TO_DATE('14-12-31', 'YY-MM-DD')
ORDER
BY FELTOLTES
Your select works like this because a CUSTOMER_ID indeed has more than one X__INSDATE, therefore the records in the result will be distinct. If you need only the first date then don't use DISTINCT and ORDER BY but try to select for MIN(X__INSDATE) and use GROUP BY CUSTOMER_ID.
SELECT DISTINCT FIRM.customer.CUSTOMER_ID,
FIRM.account_recharge.X__INSDATE FELTOLTES
Distinct is applied to both the columns together, which means you will get a distinct ROW for the set of values from the two columns. So, basically the distinct refers to all the columns in the select list.
It is equivalent to a select without distinct but a group by clause.
It means,
select distinct a, b....
is equivalent to,
select a, b...group by a, b
If you want the desired output, then CONCATENATE the columns. The distict will then work on the single concatenated resultset.
I am writing some queries with self-joins in SQL Server. When I have only one column in the SELECT clause, the query returns a certain number of rows. When I add another column, from the second instance of the table, to the SELECT clause, the results increase by 1000 rows!
How is this possible?
Thanks.
EDIT:
I have a subquery in the FROM clause, which is also a self-join on the same table.
How is this possible?
the only thing I can think of is that you have SELECT DISTINCT and the additional column makes some results distinct that weren't before the additional column.
For example I would expect the second result to have many more rows
SELECT DISTINCT First_name From Table
vs
SELECT DISTINCT First_name, Last_name From Table
But if we had the actual SQL then something else might come to mind
Simply put, I have a table with, among other things, a column for timestamps. I want to get the row with the most recent (i.e. greatest value) timestamp. Currently I'm doing this:
SELECT * FROM table ORDER BY timestamp DESC LIMIT 1
But I'd much rather do something like this:
SELECT * FROM table WHERE timestamp=max(timestamp)
However, SQLite rejects this query:
SQL error: misuse of aggregate function max()
The documentation confirms this behavior (bottom of page):
Aggregate functions may only be used in a SELECT statement.
My question is: is it possible to write a query to get the row with the greatest timestamp without ordering the select and limiting the number of returned rows to 1? This seems like it should be possible, but I guess my SQL-fu isn't up to snuff.
SELECT * from foo where timestamp = (select max(timestamp) from foo)
or, if SQLite insists on treating subselects as sets,
SELECT * from foo where timestamp in (select max(timestamp) from foo)
There are many ways to skin a cat.
If you have an Identity Column that has an auto-increment functionality, a faster query would result if you return the last record by ID, due to the indexing of the column, unless of course you wish to put an index on the timestamp column.
SELECT * FROM TABLE ORDER BY ID DESC LIMIT 1
I think I've answered this question 5 times in the past week now, but I'm too tired to find a link to one of those right now, so here it is again...
SELECT
*
FROM
table T1
LEFT OUTER JOIN table T2 ON
T2.timestamp > T1.timestamp
WHERE
T2.timestamp IS NULL
You're basically looking for the row where no other row matches that is later than it.
NOTE: As pointed out in the comments, this method will not perform as well in this kind of situation. It will usually work better (for SQL Server at least) in situations where you want the last row for each customer (as an example).
you can simply do
SELECT *, max(timestamp) FROM table
Edit:
As aggregate function can't be used like this so it gives error. I guess what SquareCog had suggested was the best thing to do
SELECT * FROM table WHERE timestamp = (select max(timestamp) from table)