Oliveiros,

I think it's time to give some more details about my task. I will start with
the begining :)

I have a "log" table which stores the dates when users send messages
with a theme from their mobile phone. This table is named
sent_messages and looks like this:

        receiver   | theme   |             date
----------------------+------------+-------------------------------
+40741775622 | CRISTI | 2007-06-04 07:44:45.406271+00
+40741775622 | CRISTI | 2007-06-04 07:45:01.788533+00
+40741775622 | CRISTI | 2007-06-04 07:45:03.764506+00
+40741775621 | LIA    | 2007-06-04 07:45:26.309215+00
+40741775621 | LIA    | 2007-06-04 07:45:28.314075+00
+40741775622 | CRISTI | 2007-06-03 07:44:00+00
+40741775622 | CRISTI | 2007-06-02 07:44:00+00
+40741775622 | CRISTI | 2007-06-01 07:44:00+00
+40741775622 | CRISTI | 2007-06-01 07:44:00+00
+40741775622 | CRISTI | 2007-06-01 07:44:00+00
+40741775622 | CRISTI | 2007-04-01 07:44:00+00
+40741775622 | CRISTI | 2007-05-01 07:44:00+00
+40741775621 | LIA    | 2007-06-03 07:44:00+00
+40741775621 | LIA    | 2007-06-03 07:44:00+00
+40741775621 | LIA    | 2007-06-02 07:44:00+00
+40741775621 | LIA    | 2007-06-02 07:44:00+00
+40741775621 | LIA    | 2007-06-01 07:44:00+00
+40741775621 | LIA    | 2007-06-01 07:44:00+00
+40741775621 | LIA    | 2007-06-01 07:44:00+00
+40741775621 | LIA    | 2007-06-01 07:44:00+00

I have also a themes table:
uid | theme
-----+--------
  6 | CRISTI
  7 | LIA

And the table named reminder_services it is used to filter
users by theme from sent_messages table. This table looks like this:

uid | theme_uid | activity_mt_amount | activity_min_days |
activity_max_months
-----+----------------+-----------------------------+---------------------------+----------------------------
   5 |              6 |                           3 |
            6 |                   1
   6 |              7 |                           4 |
   5 |                   1

The filtering should select users which sent at least activity_mt_amount
messages
with theme_uid  within activity_min_days consecutive days,
in the maximum activity_max_months months in the past.

Example:
The first row of the table reminder_services says that it should be selected
users which sent at least 3 messages with the theme_uid=6 (theme=CRISTI),
within 6 consecutive days, in the maximum 1 month in the past.


So, I created the following query:

     SELECT SUM(B.count),
                   A.theme,
                   A.receiver,
                   A.dates
        FROM ( SELECT COUNT(*),
                                   sent_messages.theme,
                                   receiver,
                                   date_trunc('day',sent_messages.date) AS
date,
                                  ARRAY(SELECT date::date + s.a FROM
generate_series(0,activity_min_days) AS s(a)) AS dates
                        FROM reminder_services, themes,sent_messages
                      WHERE themes.uid=reminder_services.theme_uid
                          AND sent_messages.theme=themes.theme
                          AND date_trunc('day',sent_messages.date) > (now()
- reminder_services.activity_max_months * INTERVAL' 1 month')
                     GROUP BY sent_messages.theme, receiver, date,
activity_min_days ) A
INNER JOIN ( SELECT COUNT(*),
                                  sent_messages.theme,
                                  receiver,
                                  date_trunc('day',sent_messages.date) AS
date,
                                  ARRAY(SELECT date::date + s.a FROM
generate_series(0,activity_min_days) AS s(a)) AS dates
                        FROM reminder_services, themes,sent_messages
                     WHERE themes.uid=reminder_services.theme_uid
                          AND sent_messages.theme=themes.theme
                          AND date_trunc('day',sent_messages.date) > (now()
- reminder_services.activity_max_months * INTERVAL' 1 month')
                     GROUP BY sent_messages.theme, receiver, date,
activity_min_days ) B
             ON A.theme=B.theme
           AND A.receiver=B.receiver
           AND B.date=ANY (A.dates)
       GROUP BY A.theme, A.receiver, A.dates;

The result of this query is:

sum | theme  |   receiver         |
dates

-----+------------+----------------------+--------------------------------------------------------------------------------
    8 | CRISTI | +40741775622 |
{2007-06-01,2007-06-02,2007-06-03,2007-06-04,2007-06-05,2007-06-06,2007-06-07}
    5 | CRISTI | +40741775622 |
{2007-06-02,2007-06-03,2007-06-04,2007-06-05,2007-06-06,2007-06-07,2007-06-08}
    4 | CRISTI | +40741775622 |
{2007-06-03,2007-06-04,2007-06-05,2007-06-06,2007-06-07,2007-06-08,2007-06-09}
    9 | CRISTI | +40741775622 |
{2007-06-04,2007-06-05,2007-06-06,2007-06-07,2007-06-08,2007-06-09,2007-06-10}
  10 | LIA       | +40741775621 |
{2007-06-01,2007-06-02,2007-06-03,2007-06-04,2007-06-05,2007-06-06}
    6 | LIA       | +40741775621 |
{2007-06-02,2007-06-03,2007-06-04,2007-06-05,2007-06-06,2007-06-07}
    4 | LIA       | +40741775621 |
{2007-06-03,2007-06-04,2007-06-05,2007-06-06,2007-06-07,2007-06-08}
    4 | LIA       | +40741775621 |
{2007-06-04,2007-06-05,2007-06-06,2007-06-07,2007-06-08,2007-06-09}
(8 rows)

The red colored values 9 and 4 are not correct, should be 3 respectively 2.


Regards,
      Loredana

Reply via email to