Hey folks,



as you may have guessed from the title, I’m an absolute beginner using
PostgreSQL and I need some help. I have a table with the columns ‘country’
(text), ‘sample site’ (text) and ‘date’ (timestamp with time zone):

country sample site date

AA a1 01.01.2000

AA a1 12.02.1999

AA a2 01.01.2000

BB b1 23.04.1984

BB b2 05.05.2000

BB b3 01.01.2000

CC c1 15.03.1998

CC c1 24.10.2002

CC c2 15.03.1998



What I want to know now is how often the different sampling sites were
tested. How do I have to write my query? I tried using something like

Select country, sample site, count (date) from samples

But the result was rubbish. I could Imagine that I have to create a loop for
this query (the table has overall more than 1 mio rows) to get a result
table with a list of sampling sites and the number of surveys for each of
them.



Help would be greatly appreciated!



Yours, Meike

On 20 August 2012 11:55, Meike Börder <boerder (AT) uni-landau (DOT) de> wrote:
Did you use a GROUP BY?

SELECT country, sample_site, count(date)
FROM samples
GROUP BY country, sample_site;

A GROUP BY is required when using aggregate functions, of which "count" is one.

--
Thom


--
Sent via pgsql-novice mailing list (pgsql-novice (AT) postgresql (DOT) org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-novice

Howdy, Meike.

Not sure if this is what you want, but
if you write

SELECT "sample site" , COUNT("date")
FROM samples
GROUP BY "sample site"

It will return the number of surveys for each one of the site.

Ain't I right?

Best,
Oliver
----- Original Message -----
From: Meike Börder
To: pgsql-novice (AT) postgresql (DOT) org
Sent: Monday, August 20, 2012 11:55 AM
Subject: [NOVICE] Postgresql-query for absolute beginner


Hey folks,



as you may have guessed from the title, I'm an absolute beginner using PostgreSQL and I need some help. I have a table with the columns 'country' (text), 'sample site' (text) and 'date' (timestamp with time zone):

country sample site date

AA a1 01.01.2000

AA a1 12.02.1999

AA a2 01.01.2000

BB b1 23.04.1984

BB b2 05.05.2000

BB b3 01.01.2000

CC c1 15.03.1998

CC c1 24.10.2002

CC c2 15.03.1998



What I want to know now is how often the different sampling sites were tested. How do I have to write my query? I tried using something like

Select country, sample site, count (date) from samples

But the result was rubbish. I could Imagine that I have to create a loop for this query (the table has overall more than 1 mio rows) to get a result table with a list of sampling sites and the number of surveys for each of them.



Help would be greatly appreciated!



Yours, Meike

thanks a lot, that worked!

cheers, Meike


-----Ursprüngliche Nachricht-----
Von: thombrown (AT) gmail (DOT) com [mailto:thombrown (AT) gmail (DOT) com] Im Auftrag von Thom
Brown
Gesendet: Montag, 20. August 2012 13:35
An: Meike Börder
Cc: pgsql-novice (AT) postgresql (DOT) org
Betreff: Re: [NOVICE] Postgresql-query for absolute beginner

On 20 August 2012 11:55, Meike Börder <boerder (AT) uni-landau (DOT) de> wrote:
Did you use a GROUP BY?

SELECT country, sample_site, count(date) FROM samples GROUP BY country,
sample_site;

A GROUP BY is required when using aggregate functions, of which "count" is
one.

--
Thom



--
Sent via pgsql-novice mailing list (pgsql-novice (AT) postgresql (DOT) org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-novice

On 20/08/12 23:38, Oliveiros d'Azevedo Cristina wrote:
count for a site should include all associated countries.

You may, or may not, be right! :-)

Anyhow, please don't top post.


Cheers,
Gavin

----- Original Message -----
From: Meike Börder
To: pgsql-novice (AT) postgresql (DOT) org
Sent: Monday, August 20, 2012 11:55 AM
Subject: [NOVICE] Postgresql-query for absolute beginner


Hey folks,



as you may have guessed from the title, I'm an absolute beginner using PostgreSQL and I need some help. I have a table with the columns 'country' (text), 'sample site' (text) and 'date' (timestamp with time zone):

country sample site date

AA a1 01.01.2000

AA a1 12.02.1999

AA a2 01.01.2000

BB b1 23.04.1984

BB b2 05.05.2000

BB b3 01.01.2000

CC c1 15.03.1998

CC c1 24.10.2002

CC c2 15.03.1998



What I want to know now is how often the different sampling sites were tested. How do I have to write my query? I tried using something like

Select country, sample site, count (date) from samples

But the result was rubbish. I could Imagine that I have to create a loop for this query (the table has overall more than 1 mio rows) to get a result table with a list of sampling sites and the number of surveys for each of them.



Help would be greatly appreciated!



Yours, Meike





You're assuming that a 'sample site' value is either unique or that the count for a site should include all associated countries.

You may, or may not, be right! :-)

Anyhow, please don't top post.

* Yeah, that wouldn't work if two sample sites on different countries happen to have the same name, but from the sample data nothing made me think so, as all sample site ID begin with the same letter as the associated country
Any way, thanks for pointin that out.

Greetings from up over,
Oliver



Cheers,
Gavin