Hi,

If you only need first/last values, and your lists are short (smaller than max_allowed_packet, given group_concat_max_len is max allowed packet) then group_concat works fine.

Window functions work /on the result set/ itself.  They are kind of like sub select over the entire result of the SELECT statement, something a subquery in the SELECT clause is unable to do.  As such window functions are not usually combined with GROUP BY.

This is because  window functions let you see the detail and the summary at the same time. They are kind of like spreadsheet expressions in that way.

For example:
SELECT date,value,last_value(value) over (partition by date order by value asc),first_value(value) over(partition by date),
sum(value) over (partition by date),
lag(value,2) over (partition by date)
From table

Shard-Query always stores the resultset in a table, so it can update the table for each window function result, which is calculated by a query.  You can look at the wf_* functions in shard-query for an implementation of each window function.  You could even add your own if you like.  If there is interest I can make that pluggable. 

You can see lots of examples in the test directory too, and s simple test data set.

--justin

Sent from my iPhone

On Oct 2, 2015, at 8:26 PM, Roberto Spadim <roberto@spadim.com.br> wrote:

hum i'm thinking about the problem
for example if we have an COUNT(*) or a SUM(value) or MAX(value) or MIN(value) or anyother functions
shouldn't be more rational allow the 'single thread' query execute and use the GROUP_CONCAT function?
ok it's not multi thread, but could the group_concat with limit 'optimize' the shard query or single queries like this?

2015-10-03 0:10 GMT-03:00 Roberto Spadim <roberto@spadim.com.br>:
nice :)

how it's "rewrite"?

2015-10-02 22:25 GMT-03:00 Justin Swanhart <greenlion@gmail.com>:
Hi,

You could use Shard-Query.  You don't need to shard the database, you can point shard-query at one database.  It supports WINDOW functions.

select 
date,
FIRST_VALUE(value) OVER (ORDER BY value ASC),
LAST_VALUE(value) OVER (ORDER BY value ASC)
FROM some_table AS a
GROUP BY date

--Justin

On Fri, Oct 2, 2015 at 6:12 PM, Roberto Spadim <roberto@spadim.com.br> wrote:
Hi guys
I have a problem and a possible solution, that i think is relative easy to develop (i didn't tested but i think the source is easy to change)

i have some queries like:


SELECT 
date,
(SELECT value FROM some_table WHERE date=a.date ORDER BY date ASC LIMIT 1) AS first,
(SELECT value FROM some_table WHERE date=a.date ORDER BY date DESC LIMIT 1) AS last
FROM some_table AS a
GROUP BY date


i want to "convert" the complex sub query, to a agregate function, and i thinking about group concat:
from mysql docs:

GROUP_CONCAT([DISTINCT] expr [,expr ...]
             [ORDER BY {unsigned_integer | col_name | expr}
                 [ASC | DESC] [,col_name ...]]
             [SEPARATOR str_val])

we have ORDER BY
if we could include a "LIMIT" clause, i could rewrite this query to:

SELECT 
date,
GROUP_CONCAT(value ORDER BY date ASC LIMIT 1) AS first,
GROUP_CONCAT(value ORDER BY date DESC LIMIT 1) AS last
FROM some_table AS a
GROUP BY date


i know that i could have stats tables / materialized views (with flexview, i already tested :) ), but i want something more "easy to use" and not "very fast", just a "feature" to solve small problems


it's a nice idea? does anyone have this "problem" and solve with other solutions?

--
Roberto Spadim

_______________________________________________
Mailing list: https://launchpad.net/~maria-discuss
Post to     : maria-discuss@lists.launchpad.net
Unsubscribe : https://launchpad.net/~maria-discuss
More help   : https://help.launchpad.net/ListHelp





--
Roberto Spadim
SPAEmpresarial - Software ERP
Eng. Automação e Controle



--
Roberto Spadim
SPAEmpresarial - Software ERP
Eng. Automação e Controle