On Mon, Nov 21, 2016 at 05:11:23PM +0100, Sandro Santilli wrote:
>
> Paul: the "histogram_cell:1" seems to be bogus in the code in any
> case, as it looks it's always set to "1", used as a flag rather than
> as an actual grid size ?
On further reading, it's only set to 1 when something "goes wron
On Mon, Nov 21, 2016 at 05:06:25PM +0100, Martin Landa wrote:
> Hi,
>
> 2016-11-21 17:00 GMT+01:00 Sandro Santilli :
> > show default_statistics_target; -- I guess 100 ?
>
> default_statistics_target | 50
Try setting them to 100, which is the default.
Then ANALYZE again, and re-run
select _pos
Hi,
2016-11-21 17:00 GMT+01:00 Sandro Santilli :
> show default_statistics_target; -- I guess 100 ?
default_statistics_target | 50
> The only thing I can think of is that the "hard deviant removal" phase
> is considering the geom in the upper-right corner as hard-deviant, as
> they are beyond th
On Mon, Nov 21, 2016 at 04:54:30PM +0100, Martin Landa wrote:
> Hi,
>
> 2016-11-21 16:38 GMT+01:00 Sandro Santilli :
> > How many rows do you have in that table ?
>
> count | 8
>
> > Did you change stat targets for the column ?
>
> No, AFAIK.
>
> > SELECT attstattarget
> >FROM pg_attribut
Hi,
2016-11-21 16:38 GMT+01:00 Sandro Santilli :
> How many rows do you have in that table ?
count | 8
> Did you change stat targets for the column ?
No, AFAIK.
> SELECT attstattarget
>FROM pg_attribute
> WHERE attrelid = 'ruian.kraje'::regclass
> AND attname = 'geom';
attstattarge
On Mon, Nov 21, 2016 at 04:28:34PM +0100, Martin Landa wrote:
> Hi,
>
> 2016-11-21 15:52 GMT+01:00 Paul Ramsey :
> > What does _postgis_stats( 'ruian.kraje', 'geom') do?
>
> select _postgis_stats( 'ruian.kraje', 'geom');
> _postgis_stats |
> {"ndims":2,"size":[1,1],"extent":{"min":[-909108,-1.233
Hi,
2016-11-21 15:52 GMT+01:00 Paul Ramsey :
> What does _postgis_stats( 'ruian.kraje', 'geom') do?
select _postgis_stats( 'ruian.kraje', 'geom');
_postgis_stats |
{"ndims":2,"size":[1,1],"extent":{"min":[-909108,-1.23343e+06],"max":[4522.92,6136.48]},"table_features":8,"sample_features":8,"not_n
What does _postgis_stats( 'ruian.kraje', 'geom') do?
On Mon, Nov 21, 2016 at 6:49 AM, Martin Landa
wrote:
> Hi,
>
> 2016-11-21 15:41 GMT+01:00 Paul Ramsey :
> > Estimated extent works by looking at the table stats. If you had some
> extra
> > data and then delete it, the stats might still reflec
Hi,
2016-11-21 15:41 GMT+01:00 Paul Ramsey :
> Estimated extent works by looking at the table stats. If you had some extra
> data and then delete it, the stats might still reflect the original extent
> of the full table. Re-running ANALYZE might refresh them.
I am afraid that this not this case.
Estimated extent works by looking at the table stats. If you had some extra
data and then delete it, the stats might still reflect the original extent
of the full table. Re-running ANALYZE might refresh them.
On Mon, Nov 21, 2016 at 5:07 AM, Martin Landa
wrote:
> Hi all,
>
> I got strange estima
Hi all,
I got strange estimated extent when working with data in S-JTSK
(EPSG:5514). ST_EstimatedExtent() returns completely wrong extent:
select st_estimated_extent('ruian', 'kraje', 'geom');
BOX(-909107.8125 -1233432.375,4522.92431640625 6136.4794921875)
The correct extent is:
select st_exten
That's right... On the target instance, the user is not exactly the same
as on the original instance.
This is due to a global settings backup problem I am trying to solve.
I will try again with the same users and tell you if it works.
Thanks for your help.
Regards,
Cedric
_
Hi Cedric,
2016-11-21 9:26 GMT+01:00 Cedric Duprez :
> Hi Giuseppe,
>
> Thanks for your answer and your help.
> Here is the output of \do @:
>
The operator signs related to PostGIS datatypes are the following ones:
> Schéma | Nom | Type de l'arg. gauche | Type de l'arg. droit | Type
> du
Hi Giuseppe,
Thanks for your answer and your help.
Here is the output of \do @:
Schéma | Nom | Type de l'arg. gauche | Type de l'arg. droit | Type
du résultat |Description
+-+---+--+--+-
14 matches
Mail list logo
