Joining sequential records in TSQL - tsql

I've got a database of messages which contain various pieces of information that come in. One such piece is a state change messages so my table looks like the following.
+-----------+--------------+---------+
| MessageId | RecievedUTC | State |
+-----------+--------------+---------+
| 1 | 1/1/2010 5pm | Off |
+-----------+--------------+---------+
| 2 | 1/2/2010 8am | Idle |
+-----------+--------------+---------+
| 3 | 1/2/2010 9am | Working |
+-----------+--------------+---------+
I'd like get a list of records which state how long I was in each state something like for a report and a maybe a pretty bar chart of how time was spent.
+---------+---------------+--------------+
| State | StartUTC | StopUTC |
+---------+---------------+--------------+
| Off | 1/1/2010 5pm | 1/2/2010 8am |
+---------+---------------+--------------+
| Idle | 1/1/2010 8am | 1/2/2010 9am |
+---------+---------------+--------------+
etc. In my mind its no harder than a join of the table with itself, offset by 1 record ordered by the RecievedUTC.
The best TSQL I could come up with is something to the effect of
SELECT m1.State, m1.RecievedUTC as StartUTC, MIN(m2.RecievedUTC) as StopUTC
FROM MessageStates as m1
INNER JOIN MessageStates as m2 ON MessageStates ON m2.RecievedUTC > m1.RecievedUTC
GROUP BY m1.MessageId, m1.State, m1.RecievedUTC
Or as a sub query to get StopUTC but both perform horribly with only 30-40k records taking almost 5 minutes to do this join.
If I wrote this in C# I would keep the track of the previous RecievedUTC and state so when I saw the next record I could combine the next RecievedUTC with it and in linear time have the data as I wanted it.

Try this:
WITH MsgStates AS
(
SELECT a.*, ROW_NUMBER() OVER(ORDER BY RecievedUTC ) RN
FROM MessageStates a
)
SELECT a.State, a.RecievedUTC StartUTC, b.RecievedUTC StartUTC
FROM MsgStates a, MsgStates b
WHERE a.rn = b.rn+1

Related

Calculate Average of Price per Items per Month in a Few Years Postgresql

I have this table inside my postgresql database,
item_code | date | price
==============================
aaaaaa.1 |2019/12/08 | 3.04
bbbbbb.b |2019/12/08 | 19.48
261893.c |2019/12/08 | 7.15
aaaaaa.1 |2019/12/17 | 4.15
bbbbbb.2 |2019/12/17 | 20
xxxxxx.5 |2019/03/12 | 3
xxxxxx.5 |2019/03/18 | 4.5
how can i calculate the average per item, per month over the year. so i get the result something like:
item_code | month | price
==============================
aaaaaa.1 | 2019/12 | 3.59
bbbbbb.2 | 2019/12 | 19.74
261893.c | 2019/12 | 7.15
xxxxxx.5 | 2019/03 | 3.75
I have tried to look and apply many alternatives but i am still not get the point, would really appreciate your help because i am new to postgresql.
I don't see how the question relates to a moving average. It seems you just want group by:
select item_code, date_trunc('month', date) as date_month, avg(price) as price
from mytable
group by item_code, date_month
This gives date_month as a date, truncated to the first day of the month - which I find more useful that the format you suggested. But it you do want that:
to_char(date, 'YYYY/MM') as date_month

Querying data with additional column that creates a number for ordering purposes

I am trying to create a "queue" system by adding an arbitrary column that creates a number based on a condition and date, to sort the importance of a row.
For example, below is the query result I pulled in Postgres:
Table: task
Result:
description | status/condition| task_created |
bla | A | 2019-12-01 07:00:00|
pikachu | A | 2019-12-01 16:32:10|
abcdef | B | 2019-12-02 18:34:22|
doremi | B | 2019-12-02 15:09:43|
lalala | A | 2019-12-03 22:10:59|
In the above, each task has a date/timestamp and status/condition applied to them. I would like to create another column that gives a number to a row where it prioritises the older tasks first, BUT if the condition is B, then we take the older task of those in B as first priority.
The expected end result (based on the example) should be:
Table1: task
description | status/condition| task_created | priority index
bla | A | 2019-12-01 07:00:00| 3
pikachu | A | 2019-12-01 16:32:10| 4
abcdef | B | 2019-12-02 18:34:22| 2
doremi | B | 2019-12-02 15:09:43| 1
lalala | A | 2019-12-03 22:10:59| 5
For priority number, 1 being most urgent to do/resolve, while 5 being the least.
How would I go about adding this additional column into the existing query? especially since there's another condition apart from just the task_created date/time.
Any help is appreciated. Many thanks!
You maybe want the Rank or Dense Rank function (depends on your needs) window functions.
If you don't need a conditional order on the status you can use this one.
SELECT *,
rank() OVER (
ORDER BY status desc, task_created
) as priority_index
FROM task
If you need a custom order based on the value of the status:
SELECT *,
rank() OVER (
ORDER BY
CASE status
WHEN 'B' THEN 1
WHEN 'A' THEN 2
WHEN 'C' THEN 3
ELSE 4
END, task_created
) as priority_index
FROM task
If you have few values this is good enough, because we can simply specify your custom order. But if you have a lot of values and the ordering information is fixed, then it should have its own table.

Stored procedure (or better way) to add a new row to existing table every day at 22:00

I will be very grateful for your advice regarding the following issue.
Given:
PostgreSQL database
Initial (basic) query
select day, Value_1, Value_2, Value_3
from table
where day=current_date
which returns a row with following columns
Day | Value_1(int) | Value_2(int) | Value 3 (int)
2019-11-14 | 10 | 10 | 14
It is needed to create a view with this starting information and add a new row every day based on the outcome of initial query executed at 22:00.
The expected outcome tomorrow at 22:01 will be
Day | Value_1 | Value_2 | Value_3
2019-11-14 | 10 | 10 | 14
2019-11-15 | N | M | P
Many thanks in advance for your time and support.

Count records between rolling date range in Tableau

I have a file with a [start] and [end] date in Tableau and would like to create a calculated field that counts number of rows on a rolling basis that occur between [start] and [end] for each [person]. This data is like so:
| Start | End | Person
|1/1/2019 |1/7/2019 | A
|1/3/2019 |1/9/2019 | A
|1/8/2019 |1/15/2019| A
|1/1/2019 |1/7/2019 | B
I'd like to create a calculated field [count] with results like so:
| Start | End | Person | Count
|1/1/2019 |1/7/2019 | A | 1
|1/3/2019 |1/9/2019 | A | 2
|1/8/2019 |1/15/2019| A | 2
|1/1/2019 |1/7/2019 | B | 1
EDITED: A good analogy for what [count] represents is: "how many videos does each person rented at the same time as of that moment?" With the 1st row for person A, count is 1, with 1 item rented. As of row 2, person A has 2 items rented. But for the 3rd row [count]= 2 since the video rented in the first row is no longer rented.

Query to combine two tables into one based on timestamp

I have three tables in Postgres. They are all about a single event (an occurrence, not "sports event"). Each table is about a specific item during the event.
table_header columns
gid, start_timestamp, end_timestamp, location, positions
table_item1 columns
gid, side, visibility, item1_timestamp
table_item2 columns
gid, position_id, name, item2_timestamp
I've tried the following query:
SELECT h.gid, h.location, h.start_timestamp, h.end_timestamp, i1.side,
i1.visibility, i2.position_id, i2.name, i2.item2_timestamp AS timestamp
FROM tablet_header AS h
LEFT OUTER JOIN table_item1 i1 on (i1.gid = h.gid)
LEFT OUTER JOIN table_item2 i2 on (i2.gid = i1.gid AND
i1.item1_timestamp = i2.item2_timestamp)
WHERE h.start_timestamp BETWEEN '2016-03-24 12:00:00'::timestamp AND now()::timestamp
The problem is that I'm losing some data from rows when item1_timestamp and item2_timestamp do not match.
So if I have in table_item1 and table_item2:
gid | item1_timestamp | side gid | item2_timestamp | name
---------------------------- -----------------------------------
1 | 17:00:00 | left 1 | 17:00:00 | charlie
1 | 17:00:05 | right 1 | 17:00:03 | frank
1 | 17:00:10 | left 1 | 17:00:06 | dee
I would want the final output to be:
gid | timestamp | side | name
-----------------------------
1 | 17:00:00 | left | charlie
1 | 17:00:03 | | frank
1 | 17:00:05 | right |
1 | 17:00:06 | | dee
1 | 17:00:10 | left |
based purely on the timestamp (and gid). Naturally I would have the header info in there too, but that's trivial.
I tried playing around with the query I posted used different JOINs and UNIONs, but I cannot seem to get it right. The one I posted gives the best results I could manage, but it's incomplete.
Side note: every minute or so there will be a new "event". So the gid will be unique to each event and the query needs to ensure that each dataset is paired with data from the same gid. Which is the reason for my i1.gid = h.gid lines. Data between different events should not be compared.
select t1.gid, t1.timestamp, t1.side, t2.name
from t1
left join t2 on t2.timestamp=t1.timestamp and t2.gid=t1.gid
union
select t1.gid, t1.timestamp, t1.side, t2.name
from t2
left join t1 on t2.timestamp=t1.timestamp and t2.gid=t1.gid