pg.ddx.io pgsql-sql@postgresql.org mailing list archive
help / color / mirror / Atom feedProblem in age on a dates interval
6+ messages / 4 participants
[nested] [flat]
* Problem in age on a dates interval
@ 2004-07-16 08:45 Luis Sousa <llsousa@ualg.pt>
0 siblings, 1 reply; 6+ messages in thread
From: Luis Sousa @ 2004-07-16 08:45 UTC (permalink / raw)
To: pgsql-sql
Hi all,
I'm using PostgreSQL 7.3.3 on i386-pc-linux-gnu, compiled by GCC gcc
(GCC) 3.3 (Debian), and I don't understand the results of the following
queries:
SELECT age('2004-05-14 16:00'::timestamp,'2004-02-18 16:00'::timestamp);
age
----------------
2 mons 25 days
SELECT '2004-02-18 16:00'::timestamp+'2 mons 25 days'::interval;
?column?
---------------------
2004-05-13 16:00:00
In this case, the age from 2004-05-14 16:00 to 2004-02-18 16:00 is 2
mons 25 days, but if I add the age to the initial date, it returns one
day less!?
SELECT age('2004-05-26 16:00'::timestamp,'2004-02-18 16:00'::timestamp);
age
---------------
3 mons 8 days
SELECT '2004-02-18 16:00'::timestamp+'3 mons 8 days'::interval;
?column?
---------------------
2004-05-26 16:00:00
Here, the age between 2004-05-26 16:00 and 2004-02-18 16:00 is 3 mons 8
days, and this interval added to the initial date gives the correct result!!
Best regards,
Luis Sousa
^ permalink raw reply [nested|flat] 6+ messages in thread
* Re: Problem in age on a dates interval
@ 2004-07-16 11:52 Theodore Petrosky <tedpet5@yahoo.com>
parent: Luis Sousa <llsousa@ualg.pt>
0 siblings, 1 reply; 6+ messages in thread
From: Theodore Petrosky @ 2004-07-16 11:52 UTC (permalink / raw)
To: Luis Sousa <llsousa@ualg.pt>; pgsql-sql
Luis,
wow.... at first I thought I had my head around a leap
year problem so I advanced your query a year....
testbed=# SELECT age('2005-05-14
16:00'::timestamp,'2005-02-18 16:00'::timestamp);
age
----------------
2 mons 24 days
(1 row)
testbed =# SELECT '2005-02-18 16:00'::timestamp +'2
mons 24 days'::interval;
?column?
---------------------
2005-05-12 16:00:00
(1 row)
I just thought I would let you know it can get
worse..... :) I don't know how daylight savings time
is playing this problem... but I didn't expect the
problem to grow like that.
Ted
--- Luis Sousa <llsousa@ualg.pt> wrote:
> Hi all,
>
> I'm using PostgreSQL 7.3.3 on i386-pc-linux-gnu,
> compiled by GCC gcc
> (GCC) 3.3 (Debian), and I don't understand the
> results of the following
> queries:
>
>
> SELECT age('2004-05-14 16:00'::timestamp,'2004-02-18
> 16:00'::timestamp);
> age
> ----------------
> 2 mons 25 days
>
> SELECT '2004-02-18 16:00'::timestamp+'2 mons 25
> days'::interval;
> ?column?
> ---------------------
> 2004-05-13 16:00:00
>
> In this case, the age from 2004-05-14 16:00 to
> 2004-02-18 16:00 is 2
> mons 25 days, but if I add the age to the initial
> date, it returns one
> day less!?
>
> SELECT age('2004-05-26 16:00'::timestamp,'2004-02-18
> 16:00'::timestamp);
> age
> ---------------
> 3 mons 8 days
>
> SELECT '2004-02-18 16:00'::timestamp+'3 mons 8
> days'::interval;
> ?column?
> ---------------------
> 2004-05-26 16:00:00
>
> Here, the age between 2004-05-26 16:00 and
> 2004-02-18 16:00 is 3 mons 8
> days, and this interval added to the initial date
> gives the correct result!!
>
__________________________________
Do you Yahoo!?
Yahoo! Mail - 50x more storage than other providers!
http://promotions.yahoo.com/new_mail
^ permalink raw reply [nested|flat] 6+ messages in thread
* Re: Problem in age on a dates interval
@ 2004-07-16 14:34 Tom Lane <tgl@sss.pgh.pa.us>
parent: Theodore Petrosky <tedpet5@yahoo.com>
0 siblings, 1 reply; 6+ messages in thread
From: Tom Lane @ 2004-07-16 14:34 UTC (permalink / raw)
To: Theodore Petrosky <tedpet5@yahoo.com>; +Cc: Luis Sousa <llsousa@ualg.pt>; pgsql-sql
Theodore Petrosky <tedpet5@yahoo.com> writes:
> wow.... at first I thought I had my head around a leap
> year problem so I advanced your query a year....
I think what's going on here is a difference of interpretation about
whether an "M months D days" interval means to add the months first
or the days first. For instance
2005-02-18 plus 2 months = 2005-04-18, plus 24 days = 2005-05-12
2005-02-18 plus 24 days = 2005-03-14, plus 2 months = 2005-05-14
The timestamp-plus-interval operator is evidently doing addition the
first way, but it looks like age() is calculating the difference in a
way that implicitly corresponds to the second way.
I have some vague recollection that this has come up before, but
I don't recall whether we concluded that age() needs to be changed
or not. In any case it's not risen to the top of anyone's to-do list,
because I see that age() still acts this way in CVS tip.
regards, tom lane
^ permalink raw reply [nested|flat] 6+ messages in thread
* Re: Problem in age on a dates interval
@ 2004-07-19 09:00 Luis Sousa <llsousa@ualg.pt>
parent: Tom Lane <tgl@sss.pgh.pa.us>
0 siblings, 1 reply; 6+ messages in thread
From: Luis Sousa @ 2004-07-19 09:00 UTC (permalink / raw)
To: Tom Lane <tgl@sss.pgh.pa.us>; +Cc: Theodore Petrosky <tedpet5@yahoo.com>; pgsql-sql
I worked around this problem returning the difference between the two
dates, using extract doy from both.
Anyway, this will cause a bug on my code when changing the year. Any ideas?
Best regards,
Luis Sousa
Tom Lane wrote:
>Theodore Petrosky <tedpet5@yahoo.com> writes:
>
>
>>wow.... at first I thought I had my head around a leap
>>year problem so I advanced your query a year....
>>
>>
>
>I think what's going on here is a difference of interpretation about
>whether an "M months D days" interval means to add the months first
>or the days first. For instance
>
>2005-02-18 plus 2 months = 2005-04-18, plus 24 days = 2005-05-12
>
>2005-02-18 plus 24 days = 2005-03-14, plus 2 months = 2005-05-14
>
>The timestamp-plus-interval operator is evidently doing addition the
>first way, but it looks like age() is calculating the difference in a
>way that implicitly corresponds to the second way.
>
>I have some vague recollection that this has come up before, but
>I don't recall whether we concluded that age() needs to be changed
>or not. In any case it's not risen to the top of anyone's to-do list,
>because I see that age() still acts this way in CVS tip.
>
> regards, tom lane
>
>
>
>
^ permalink raw reply [nested|flat] 6+ messages in thread
* Re: Problem in age on a dates interval
@ 2004-07-19 10:46 Alexander M. Pravking <fduch@antar.bryansk.ru>
parent: Luis Sousa <llsousa@ualg.pt>
0 siblings, 1 reply; 6+ messages in thread
From: Alexander M. Pravking @ 2004-07-19 10:46 UTC (permalink / raw)
To: Luis Sousa <llsousa@ualg.pt>; +Cc: Tom Lane <tgl@sss.pgh.pa.us>; Theodore Petrosky <tedpet5@yahoo.com>; pgsql-sql
On Mon, Jul 19, 2004 at 10:00:50AM +0100, Luis Sousa wrote:
> I worked around this problem returning the difference between the two
> dates, using extract doy from both.
> Anyway, this will cause a bug on my code when changing the year. Any ideas?
Why don't you use the minus operator?
SELECT '2004-05-14 16:00'::timestamp - '2004-02-18 16:00'::timestamp;
?column?
----------
86 days
Or, if you need the age just in days:
SELECT extract(day from '2004-05-14 16:00'::timestamp - '2004-02-18 16:00'::timestamp);
date_part
-----------
86
or
SELECT '2004-05-14 16:00'::date - '2004-02-18 16:00'::date;
?column?
----------
86
Note that '2004-05-14 16:00'::date is actually '2004-05-14 00:00'::date,
so the last two are not always equal.
> Tom Lane wrote:
>
> >Theodore Petrosky <tedpet5@yahoo.com> writes:
> >
> >
> >>wow.... at first I thought I had my head around a leap
> >>year problem so I advanced your query a year....
> >>
> >>
> >
> >I think what's going on here is a difference of interpretation about
> >whether an "M months D days" interval means to add the months first
> >or the days first. For instance
> >
> >2005-02-18 plus 2 months = 2005-04-18, plus 24 days = 2005-05-12
> >
> >2005-02-18 plus 24 days = 2005-03-14, plus 2 months = 2005-05-14
> >
> >The timestamp-plus-interval operator is evidently doing addition the
> >first way, but it looks like age() is calculating the difference in a
> >way that implicitly corresponds to the second way.
> >
> >I have some vague recollection that this has come up before, but
> >I don't recall whether we concluded that age() needs to be changed
> >or not. In any case it's not risen to the top of anyone's to-do list,
> >because I see that age() still acts this way in CVS tip.
> >
> > regards, tom lane
--
Fduch M. Pravking
^ permalink raw reply [nested|flat] 6+ messages in thread
* Re: Problem in age on a dates interval
@ 2004-07-20 08:17 Luis Sousa <llsousa@ualg.pt>
parent: Alexander M. Pravking <fduch@antar.bryansk.ru>
0 siblings, 0 replies; 6+ messages in thread
From: Luis Sousa @ 2004-07-20 08:17 UTC (permalink / raw)
To: Alexander M. Pravking <fduch@antar.bryansk.ru>; +Cc: Tom Lane <tgl@sss.pgh.pa.us>; Theodore Petrosky <tedpet5@yahoo.com>; pgsql-sql
Yes, that's a much more clever solution than the one I used.
Thanks
Best regards,
Luis Sousa
Alexander M. Pravking wrote:
>On Mon, Jul 19, 2004 at 10:00:50AM +0100, Luis Sousa wrote:
>
>
>>I worked around this problem returning the difference between the two
>>dates, using extract doy from both.
>>Anyway, this will cause a bug on my code when changing the year. Any ideas?
>>
>>
>
>Why don't you use the minus operator?
>
>SELECT '2004-05-14 16:00'::timestamp - '2004-02-18 16:00'::timestamp;
> ?column?
>----------
> 86 days
>
>Or, if you need the age just in days:
>
>SELECT extract(day from '2004-05-14 16:00'::timestamp - '2004-02-18 16:00'::timestamp);
> date_part
>-----------
> 86
>
>or
>
>SELECT '2004-05-14 16:00'::date - '2004-02-18 16:00'::date;
> ?column?
>----------
> 86
>
>Note that '2004-05-14 16:00'::date is actually '2004-05-14 00:00'::date,
>so the last two are not always equal.
>
>
>
>
>>Tom Lane wrote:
>>
>>
>>
>>>Theodore Petrosky <tedpet5@yahoo.com> writes:
>>>
>>>
>>>
>>>
>>>>wow.... at first I thought I had my head around a leap
>>>>year problem so I advanced your query a year....
>>>>
>>>>
>>>>
>>>>
>>>I think what's going on here is a difference of interpretation about
>>>whether an "M months D days" interval means to add the months first
>>>or the days first. For instance
>>>
>>>2005-02-18 plus 2 months = 2005-04-18, plus 24 days = 2005-05-12
>>>
>>>2005-02-18 plus 24 days = 2005-03-14, plus 2 months = 2005-05-14
>>>
>>>The timestamp-plus-interval operator is evidently doing addition the
>>>first way, but it looks like age() is calculating the difference in a
>>>way that implicitly corresponds to the second way.
>>>
>>>I have some vague recollection that this has come up before, but
>>>I don't recall whether we concluded that age() needs to be changed
>>>or not. In any case it's not risen to the top of anyone's to-do list,
>>>because I see that age() still acts this way in CVS tip.
>>>
>>> regards, tom lane
>>>
>>>
>
>
>
^ permalink raw reply [nested|flat] 6+ messages in thread
end of thread, other threads:[~2004-07-20 08:17 UTC | newest]
Thread overview: 6+ messages (download: mbox mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
2004-07-16 08:45 Problem in age on a dates interval Luis Sousa <llsousa@ualg.pt>
2004-07-16 11:52 ` Theodore Petrosky <tedpet5@yahoo.com>
2004-07-16 14:34 ` Tom Lane <tgl@sss.pgh.pa.us>
2004-07-19 09:00 ` Luis Sousa <llsousa@ualg.pt>
2004-07-19 10:46 ` Alexander M. Pravking <fduch@antar.bryansk.ru>
2004-07-20 08:17 ` Luis Sousa <llsousa@ualg.pt>
This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox