kdb+ equivalent of SQL's rank() and dense_rank() - kdb

Any one every have to simulate the result of SQL's rank(), dense_rank(), and row_number(), in kdb+? Here is some SQL to demonstrate the features. If anyone has a specific solution below, perhaps I could work on generalising it to support multiple partition and order by columns -- and post back on this site.
CREATE TABLE student(course VARCHAR(10), mark int, name varchar(10));
INSERT INTO student VALUES
('Maths', 60, 'Thulile'),
('Maths', 60, 'Pritha'),
('Maths', 70, 'Voitto'),
('Maths', 55, 'Chun'),
('Biology', 60, 'Bilal'),
('Biology', 70, 'Roger');
SELECT
RANK() OVER (PARTITION BY course ORDER BY mark DESC) AS rank,
DENSE_RANK() OVER (PARTITION BY course ORDER BY mark DESC) AS dense_rank,
ROW_NUMBER() OVER (PARTITION BY course ORDER BY mark DESC) AS row_num,
course, mark, name
FROM student ORDER BY course, mark DESC;
+------+------------+---------+---------+------+---------+
| rank | dense_rank | row_num | course | mark | name |
+------+------------+---------+---------+------+---------+
| 1 | 1 | 1 | Biology | 70 | Roger |
| 2 | 2 | 2 | Biology | 60 | Bilal |
| 1 | 1 | 1 | Maths | 70 | Voitto |
| 2 | 2 | 2 | Maths | 60 | Thulile |
| 2 | 2 | 3 | Maths | 60 | Pritha |
| 4 | 3 | 4 | Maths | 55 | Chun |
+------+------------+---------+---------+------+---------+
Here is some kdb+ to generate the equivalent student table:
student:([] course:`Maths`Maths`Maths`Maths`Biology`Biology;
mark:60 60 70 55 60 70;
name:`Thulile`Pritha`Voitto`Chun`Bilal`Roger)
Thank you!

If you sort the table initially by course and mark:
student:`course xasc `mark xdesc ([] course:`Maths`Maths`Maths`Maths`Biology`Biology;mark:60 60 70 55 60 70;name:`Thulile`Pritha`Voitto`Chun`Bilal`Roger)
course mark name
--------------------
Biology 70 Roger
Biology 60 Bilal
Maths 70 Voitto
Maths 60 Thulile
Maths 60 Pritha
Maths 55 Chun
Then you can use something like the below to achieve your output:
update rank_sql:first row_num by course,mark from update dense_rank:1+where count each (where differ mark)cut mark,row_num:1+rank i by course from student
course mark name dense_rank row_num rank_sql
------------------------------------------------
Biology 70 Roger 1 1 1
Biology 60 Bilal 2 2 2
Maths 70 Voitto 1 1 1
Maths 60 Thulile 2 2 2
Maths 60 Pritha 2 3 2
Maths 55 Chun 3 4 4
This solution uses rank and the virtual index column if you would like to read up further on these.

For table ordered by target columns:
q) dense_sql:{sums differ x}
q) rank_sql:{raze #'[(1_deltas b),1;b:1+where differ x]}
q) row_sql:{1+til count x}
q) student:`course xasc `mark xdesc ([] course:`Maths`Maths`Maths`Maths`Biology`Biology;mark:60 60 70 55 60 70;name:`Thulile`Pritha`Voitto`Chun`Bilal`Roger)
q)update row_num:row_sql mark,rank_s:rank_sql mark,dense_s:dense_sql mark by course from student

I can think of this as of now:
Note: The rank function in kdb works on asc list, so I created below functions.
I would not xdesc the table, as I can just use the vector column and desc it
q)denseF
{((desc distinct x)?x)+1}
q)rankF
{((desc x)?x)+1}
q)update dense_rank:denseF mark,rank_rank:rankF mark,row_num:1+rank i by course from student
course
mark name
dense_rank
rank_rank
row_num
Maths
60 Thulile
2
2
1
Maths
60 Pritha
2
2
2
Maths
70 Voitto
1
1
3
Maths
55 Chun
3
4
4
Biology
60 Bilal
2
2
1
Biology
70 Roger
1
1
2

Related

Unpivot data in PostgreSQL

I have a table in PostgreSQL with the below values,
empid hyderabad bangalore mumbai chennai
1 20 30 40 50
2 10 20 30 40
And my output should be like below
empid city nos
1 hyderabad 20
1 bangalore 30
1 mumbai 40
1 chennai 50
2 hyderabad 10
2 bangalore 20
2 mumbai 30
2 chennai 40
How can I do this unpivot in PostgreSQL?
You can use a lateral join:
select t.empid, x.city, x.nos
from the_table t
cross join lateral (
values
('hyderabad', t.hyderabad),
('bangalore', t.bangalore),
('mumbai', t.mumbai),
('chennai', t.chennai)
) as x(city, nos)
order by t.empid, x.city;
Or this one: simpler to read- and real plain SQL ...
WITH
input(empid,hyderabad,bangalore,mumbai,chennai) AS (
SELECT 1,20,30,40,50
UNION ALL SELECT 2,10,20,30,40
)
,
i(i) AS (
SELECT 1
UNION ALL SELECT 2
UNION ALL SELECT 3
UNION ALL SELECT 4
)
SELECT
empid
, CASE i
WHEN 1 THEN 'hyderabad'
WHEN 2 THEN 'bangalore'
WHEN 3 THEN 'mumbai'
WHEN 4 THEN 'chennai'
ELSE 'unknown'
END AS city
, CASE i
WHEN 1 THEN hyderabad
WHEN 2 THEN bangalore
WHEN 3 THEN mumbai
WHEN 4 THEN chennai
ELSE NULL::INT
END AS city
FROM input CROSS JOIN i
ORDER BY empid,i;
-- out empid | city | city
-- out -------+-----------+------
-- out 1 | hyderabad | 20
-- out 1 | bangalore | 30
-- out 1 | mumbai | 40
-- out 1 | chennai | 50
-- out 2 | hyderabad | 10
-- out 2 | bangalore | 20
-- out 2 | mumbai | 30
-- out 2 | chennai | 40

PostgresQL for each row, generate new rows and merge

I have a table called example that looks as follows:
ID | MIN | MAX |
1 | 1 | 5 |
2 | 34 | 38 |
I need to take each ID and loop from it's min to max, incrementing by 2 and thus get the following WITHOUT using INSERT statements, thus in a SELECT:
ID | INDEX | VALUE
1 | 1 | 1
1 | 2 | 3
1 | 3 | 5
2 | 1 | 34
2 | 2 | 36
2 | 3 | 38
Any ideas of how to do this?
The set-returning function generate_series does exactly that:
SELECT
id,
generate_series(1, (max-min)/2+1) AS index,
generate_series(min, max, 2) AS value
FROM
example;
(online demo)
The index can alternatively be generated with RANK() (example, see also #a_horse_­with_­no_­name's answer) if you don't want to rely on the parallel sets.
Use generate_series() to generate the numbers and a window function to calculate the index:
select e.id,
row_number() over (partition by e.id order by g.value) as index,
g.value
from example e
cross join generate_series(e.min, e.max, 2) as g(value);

How would you create a group identifier based on one column, but sorted by another?

I am attempting to create column Group via T-SQL.
If a cluster of accounts are in a row, consider that as one group. if the account is seen again lower in the list (cluster or not), then consider it a new group. This seems straight forward, but I cannot seem to see the solution... Below there are three clusters of account 3456, each having a different group number (Group 1,4, and 6)
+-------+---------+------+
| Group | Account | Sort |
+-------+---------+------+
| 1 | 3456 | 1 |
| 1 | 3456 | 2 |
| 2 | 9878 | 3 |
| 3 | 5679 | 4 |
| 4 | 3456 | 5 |
| 4 | 3456 | 6 |
| 4 | 3456 | 7 |
| 5 | 1295 | 8 |
| 6 | 3456 | 9 |
+-------+---------+------+
UPDATE: I left this out of the original requirements, but a cluster of accounts could have more than two accounts. I updated the example data to include this scenario.
Here's how I'd do it:
--Sample Data
DECLARE #table TABLE (Account INT, Sort INT);
INSERT #table
VALUES (3456,1),(3456,2),(9878,3),(5679,4),(3456,5),(3456,6),(1295,7),(3456,8);
--Solution
SELECT [Group] = DENSE_RANK() OVER (ORDER BY grouper.groupID), grouper.Account, grouper.Sort
FROM
(
SELECT t.*, groupID = ROW_NUMBER() OVER (ORDER BY t.sort) +
CASE t.Account WHEN LEAD(t.Account,1) OVER (ORDER BY t.sort) THEN 1 ELSE 0 END
FROM #table AS t
) AS grouper;
Results:
Group Account Sort
------- ----------- -----------
1 3456 1
1 3456 2
2 9878 3
3 5679 4
4 3456 5
4 3456 6
5 1295 7
6 3456 8
Update based on OPs comment below (20190508)
I spent a couple days banging my head on how to handle groups of three or more; it was surprisingly difficult but what I came up with handles bigger clusters and is way better than my first answer. I updated the sample data to include bigger clusters.
Note that I include a UNIQUE constraint for the sort column - this creates a unique index. You don't need the constraint for this solution to work but, having an index on that column (clustered, nonclustered unique or just nonclustered) will improve the performance dramatically.
--Sample Data
DECLARE #table TABLE (Account INT, Sort INT UNIQUE);
INSERT #table
VALUES (3456,1),(3456,2),(9878,3),(5679,4),(3456,5),(3456,6),(1295,7),(1295,8),(1295,9),(1295,10),(3456,11);
-- Better solution
WITH Groups AS
(
SELECT t.*, Grouper =
CASE t.Account WHEN LAG(t.Account,1,t.Account) OVER (ORDER BY t.Sort) THEN 0 ELSE 1 END
FROM #table AS t
)
SELECT [Group] = SUM(sg.Grouper) OVER (ORDER BY sg.Sort)+1, sg.Account, sg.Sort
FROM Groups AS sg;
Results:
Group Account Sort
----------- ----------- -----------
1 3456 1
1 3456 2
2 9878 3
3 5679 4
4 3456 5
4 3456 6
5 1295 7
5 1295 8
5 1295 9
5 1295 10
6 3456 11

de-aggregate for table columns in Greenplum

I am using Greenplum, and I have data like:
id | val
----+-----
12 | 12
12 | 23
12 | 34
13 | 23
13 | 34
13 | 45
(6 rows)
somehow I want the result like:
id | step
----+-----
12 | 12
12 | 11
12 | 11
13 | 23
13 | 11
13 | 11
(6 rows)
How it comes:
First there should be a Window function, which execute a de-aggreagte function based on partition by id
the column val is cumulative value, and what I want to get is the step values.
Maybe I can do it like:
select deagg(val) over (partition by id) from table_name;
So I need the deagg function.
Thanks for your help!
P.S and Greenplum is based on postgresql v8.2
You can just use the LAG function:
SELECT id,
val - lag(val, 1, 0) over (partition BY id ORDER BY val) as step
FROM yourTable
Note carefully that lag() has three parameters. The first is the column for which to find the lag, the second indicates to look at the previous record, and the third will cause lag to return a default value of zero.
Here is a table showing the table this query would generate:
id | val | lag(val, 1, 0) | val - lag(val, 1, 0)
----+-----+----------------+----------------------
12 | 12 | 0 | 12
12 | 23 | 12 | 11
12 | 34 | 23 | 11
13 | 23 | 0 | 23
13 | 34 | 23 | 11
13 | 45 | 34 | 11
Second note: This answer assumes that you want to compute your rolling difference in order of val ascending. If you want a different order you can change the ORDER BY clause of the partition.
val seems to be a cumulative sum. You can "unaggregate" it by subtracting the previous val from the current val, e.g., by using the lag function. Just note you'll have to treat the first value in each group specially, as lag will return null:
SELECT id, val - COALESCE(LAG(val) OVER (PARTITION BY id ORDER BY val), 0) AS val
FROM mytable;

postgresql Subselect Aggregate in larger query

I'm working with a gigantic dataset of individuals with demographic information and action tracking. I am trying to get the percentage of people who committed an action, which is simple, but also trying to get average ages of people who fit in a specific subgroup of the original SELECT. The CASE WHEN line works fine alone, and the subquery runs fine in it's own query but I cannot seem to get it integrated into this query as a subquery, it gives me a syntax error on the CASE WHEN statement. Here's a slightly anonymized version of the query. Any help would be VERY appreciated.
SELECT
AVG(ageagg)
FROM
(
SELECT
age AS ageagg
FROM
agetable
WHERE
age>30
AND action_taken=1) AvgAge_30Action,
COUNT(
CASE
WHEN action_taken=1
AND age> 30
THEN 1
ELSE 0 NULL) / COUNT(
CASE
WHEN age>30) AS Over_30_Action
FROM
agetable
WHERE
website_type=3
If I've interpreted your intent correctly, you wish to compute the following:
1) the number of people over the age of 30 that took a specific action as a percentage of the total number of people over the age of 30
2) the average age of the people over the age of 30 that took a specific action
Assuming my interpretation is correct, this query might work for you:
SELECT
100 * over_30_action / over_30_total AS percentage_of_over_30_took_action,
average_age_of_over_30_took_action
FROM (
SELECT
SUM(CASE WHEN action_taken=1 THEN 1 ELSE 0 END) AS over_30_action,
COUNT(*) AS over_30_total,
AVG(CASE WHEN action_taken=1 THEN age ELSE NULL END)
AS average_age_of_over_30_took_action
FROM agetable
WHERE website_type=3 AND age>30
) aggregated;
I created a dummy table and populated it with the following data.
postgres=# select * from agetable order by website_type, action_taken, age;
age | action_taken | website_type
-----+--------------+--------------
33 | 1 | 1
32 | 1 | 2
28 | 1 | 3
29 | 1 | 3
32 | 1 | 3
33 | 1 | 3
34 | 1 | 3
32 | 2 | 3
32 | 3 | 3
33 | 4 | 3
34 | 5 | 3
33 | 6 | 3
34 | 7 | 3
35 | 8 | 3
(14 rows)
Of the 14 rows, 4 rows (the first four in this listing) have either the wrong website_type or have age below 30. Of the ten remaining rows, you can see that 3 of them have an action_taken of 1. So, the query should determine that 30% of folks over the age of 30 took a particular action, and the average age among that particular population should be 33 (ages 32, 33, and 34). The results of the query I posted:
percentage_of_over_30_took_action | average_age_of_over_30_took_action
-----------------------------------+------------------------------------
30 | 33.0000000000000000
(1 row)
Again, all of this is predicated upon my interpretation of your intent actually being accurate. This is of course based on a highly contrived data set, but hopefully it's enough of a functional signpost to get you on the right path.