>SELECT (ST_SummaryStats(..)).* FROM ...
>
>Should show you "labels" to better understand.
Yes, thx, now it report me this:
count | sum | mean | stddev | min | max
-------+-----+------+--------+-----+-----
10000 | 0 | 0 | 0 | 0 | 0
10000 | 0 | 0 | 0 | 0 | 0
10000 | 0 | 0 | 0 | 0 | 0
10000 | 0 | 0 | 0 | 0 | 0
10000 | 0 | 0 | 0 | 0 | 0
10000 | 0 | 0 | 0 | 0 | 0
10000 | 0 | 0 | 0 | 0 | 0
10000 | 0 | 0 | 0 | 0 | 0
10000 | 0 | 0 | 0 | 0 | 0
10000 | 0 | 0 | 0 | 0 | 0
.....
>I guess you could run aggregates on the single columns of the stats, like:
>
>WITH stats AS ( SELECT (ST_SummaryStats(..)).* FROM ... )
>SELECT sum(count) as count, sum(sum) ... FROM stats;
Thx , strk.
Running this query
WITH stats AS (
SELECT
(ST_SummaryStats(rast)).*
FROM
my_schema.my_table
)
SELECT
sum(count) as count,
sum(sum) as sum,
avg(mean) as mean,
avg(stddev) as stddev,
min(min) as min,
max(max) as max
FROM
stats
;
I have the report I need.
count | sum | mean | stddev | min | max
-----------+-----------+------------------+------------------+-----+-----
164020000 | 620088072 | 3.78056378490428 | 2.01499851432122 | 0 | 245
(1 riga)
>I guess it could be useful to define an aggregate taking
>"SummaryStats" type in input so you could do something like:
>
>SELECT (ST_SummaryStatsAgg(ST_SummaryStats(xx))).* FROM myrast;
I guess it should better.
--
-----------------
Andrea Peri
. . . . . . . . .
qwerty àèìòù
-----------------
_______________________________________________
postgis-users mailing list
postgis-users@postgis.refractions.net
http://postgis.refractions.net/mailman/listinfo/postgis-users
