Re: pgadmin error

From: Dave Page <dpage(at)pgadmin(dot)org>
To: Charlie Holleran <charles(dot)holleran(at)shiftenergy(dot)com>
Cc: pgadmin-hackers <pgadmin-hackers(at)postgresql(dot)org>
Subject: Re: pgadmin error
Date: 2012-05-18 03:55:05
Message-ID: CA+OCxoxRUD=fdJrpbdr8CuX2A-q+agXbwnMGhX4c4Sn-01AS=Q@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgadmin-hackers

Thanks Charlie. What happens with psql, per my question below?

On Thu, May 17, 2012 at 10:24 PM, Charlie Holleran
<charles(dot)holleran(at)shiftenergy(dot)com> wrote:
> INSERT INTO tbla(col1, col2, date_observed, rt_value)
> VALUES
> ( -1465867942,-3248196,'2012-05-09 23:57:00-03',8536388),
> ( -1465867942,-3248196,'2012-05-09 23:58:00-03',8536396),
> ( -1465867942,-3248196,'2012-05-09 23:59:00-03',8536404);
>
>
>
>
> --------------------------------------------------
> From: "Dave Page" <dpage(at)pgadmin(dot)org>
> Sent: Thursday, May 17, 2012 5:45 PM
>
> To: "Charlie Holleran" <charles(dot)holleran(at)shiftenergy(dot)com>
> Cc: <pgadmin-hackers(at)postgresql(dot)org>
> Subject: Re: [pgadmin-hackers] pgadmin error
>
>> HI
>>
>> Can I get the raw data for the test case please (ideally as a set of
>> INSERT statements)?
>>
>> Also, what happens if you INSERT/SELECT the same data using psql?
>>
>> On Wed, May 16, 2012 at 10:24 PM, Charlie Holleran
>> <charles(dot)holleran(at)shiftenergy(dot)com> wrote:
>>>
>>> Hi Dave
>>>
>>> Tried some more queries to qualify the rounding problem accessing data in
>>> pgadmin.  Not quite what was expected.  Hmmm...
>>>
>>> Edit view brings back the values such as 8.53639e+06.
>>>
>>> What has been stored in the table are numbers in increasing value.
>>>
>>> <test_query>
>>>
>>> SELECT col1, col2, date_observed,
>>>  rt_value v0, rt_value::TEXT v1, to_char(rt_value, '999999999') v2,
>>>  rt_value::NUMERIC v3, rt_value::NUMERIC::TEXT v4,
>>> to_char(rt_value::NUMERIC, '999999999') v5,
>>>  rt_value::double precision v6, rt_value::double precision::TEXT v7,
>>> to_char(rt_value::double precision, '999999999') v8
>>> FROM tbla
>>> WHERE (col1,col2) = ( (-1465867942,-3248196) )
>>> AND date_observed >= '9-May-2012 23:57'
>>> AND date_observed < '10-May-2012'
>>> ORDER BY col1, col2, date_observed
>>> ;
>>>
>>> </test_query>
>>>
>>> <results>
>>>
>>> -1465867942;-3248196;"2012-05-09 23:57:00-03";8.53639e+06;"8.53639e+06";"
>>> 8536388";8536390;"8536390";"   8536390";8536388;"8536388";"   8536388"
>>> -1465867942;-3248196;"2012-05-09 23:58:00-03";8.5364e+06;"8.5364e+06";"
>>> 8536396";8536400;"8536400";"   8536400";8536396;"8536396";"   8536396"
>>> -1465867942;-3248196;"2012-05-09 23:59:00-03";8.5364e+06;"8.5364e+06";"
>>> 8536404";8536400;"8536400";"   8536400";8536404;"8536404";"   8536404"
>>>
>>> </results>
>>>
>>> <table_snippet>
>>>
>>> CREATE TABLE tbldaq_adb
>>> (
>>>  id bigserial NOT NULL,
>>>  col1 integer NOT NULL,
>>>  col2 integer NOT NULL,
>>>  date_observed timestamp with time zone NOT NULL,
>>>  rt_value real NOT NULL,
>>> );
>>>
>>> </table_snippet>
>>>
>>> Charlie
>>>
>>>
>>>
>>> --------------------------------------------------
>>> From: "Dave Page" <dpage(at)pgadmin(dot)org>
>>> Sent: Wednesday, May 16, 2012 2:51 PM
>>> To: "Charlie Holleran" <charles(dot)holleran(at)shiftenergy(dot)com>
>>> Cc: <pgadmin-hackers(at)postgresql(dot)org>
>>> Subject: Re: [pgadmin-hackers] pgadmin error
>>>
>>>
>>>> On Wed, May 16, 2012 at 10:04 AM, Charlie Holleran
>>>> <charles(dot)holleran(at)shiftenergy(dot)com> wrote:
>>>>>
>>>>>
>>>>> Is there a way to configure the JDBC driver that pgadmin uses to access
>>>>> the
>>>>> db's?  I'm using pg's postgresql-8.4-701.jdbc4.jar in my code and I get
>>>>> the
>>>>> correct values from the db.  But pgadmin does not get the correct
>>>>> values
>>>>> back.  pgadmin is rounding real's to 6 significant digits, independent
>>>>> of
>>>>> the magnitude of the value.  real's obviously have an issue with
>>>>> resolution
>>>>> as the magnitude of the number increases.  But this pgadmin issue is
>>>>> not
>>>>> magnitude dependent.  Second the real datatype issue loses resolution
>>>>> in
>>>>> power of 2's.  pgadmin is losing in the power of 10's.  It seems like a
>>>>> string parsing issue in the jdbc that only parses the first 6 digits.
>>>>>
>>>>> So I'd like to configure pgadmin to use the same jdbc driver as my
>>>>> code.
>>>>> Which works.  Where can I set the driver for pgadmin?
>>>>
>>>>
>>>>
>>>> pgAdmin isn't a Java app. It's written in C++ and uses libpq to access
>>>> the server.
>>>>
>>>> Where in pgAdmin do you see the rounding? (query tool, edit grid
>>>> or...). Can you provide a self contained test case?
>>>>
>>>>
>>>> --
>>>> Dave Page
>>>> Blog: http://pgsnake.blogspot.com
>>>> Twitter: @pgsnake
>>>>
>>>> EnterpriseDB UK: http://www.enterprisedb.com
>>>> The Enterprise PostgreSQL Company
>>>>
>>>
>>
>>
>>
>> --
>> Dave Page
>> Blog: http://pgsnake.blogspot.com
>> Twitter: @pgsnake
>>
>> EnterpriseDB UK: http://www.enterprisedb.com
>> The Enterprise PostgreSQL Company
>>
>

--
Dave Page
Blog: http://pgsnake.blogspot.com
Twitter: @pgsnake

EnterpriseDB UK: http://www.enterprisedb.com
The Enterprise PostgreSQL Company

In response to

Browse pgadmin-hackers by date

  From Date Subject
Next Message Guillaume Lelarge 2012-05-19 11:49:00 pgAdmin III commit: Update the TODO list
Previous Message Dave Page 2012-05-17 20:45:50 Re: pgadmin error