I forgot to specify the unique key for the table to be updated
in the first attempt at INSERT OR REPLACE.
Please try this...
insert or replace into A3Test115_J(
PATIENT_ID, ---<---
ENTRY_ID_E2,
READ_CODE_E2,
TERM_TEXT_E2,
...
NUMERIC_VALUE_E15
)
select
t1.PATIENT_ID,
--- Joe Wilson <[EMAIL PROTECTED]> wrote:
> I think this technique might be more compact and efficient than the update:
I think my suggestion may result in duplicate rows.
If the INSERT or REPLACE doesn't work, you might try:
1. BEGIN
2. dump the joined select into a temp table
3. delete the
Thanks, will have a look at that.
It definitely looks better.
RBS
-Original Message-
From: Joe Wilson [mailto:[EMAIL PROTECTED]
Sent: 28 January 2007 20:10
To: sqlite-users@sqlite.org
Subject: Re: [sqlite] Limit statement size?
I think this technique might be more compact and
I think this technique might be more compact and efficient than the update:
insert or replace into A3Test115_J(
ENTRY_ID_E2,
READ_CODE_E2,
TERM_TEXT_E2,
...
NUMERIC_VALUE_E15
)
select
g2.ENTRY_ID,
g2.READ_CODE,
g2.TERM_TEXT,
...
g15.NUMERIC_VALUE
from
A3Test115_J t1,
> The UPDATE statement in the original example is a way to
> compensate for SQLite's inability to perform a join on
> another table before doing an UPDATE.
Yes, I remember trying to do this at first via joins, but I couldn't
get that to work and ended up with the current solution.
Still, it
--- Fred Williams <[EMAIL PROTECTED]> wrote:
> Wow! Talk about obfuscated code! I didn't even try to dig deeper than
> a quick scan, but could this abomination be broken into multiple update
> queries? On the surface it looks like each "group" is unique. If so,
> wouldn't a transaction with
I don't think normalization comes into play here.
These are a whole of manipulations to transpose a table.
My normal base tables are all normalized fine.
RBS
-Original Message-
From: Clay Dowling [mailto:[EMAIL PROTECTED]
Sent: 28 January 2007 19:01
To: sqlite-users@sqlite.org
Subject:
You have any way to normalize that original source table? It's full of
extents, which is the first sign of badness in a table design and
assured of complicating your life.
Clay
RB Smissaert wrote:
> Yes, I agree it looks messy, but I used to do this in steps and after advice
> I think from Igor
Yes, I agree it looks messy, but I used to do this in steps and after advice
I think from Igor Tandenik I lumped it all together and run it in one go,
which is a lot faster.
RBS
-Original Message-
From: Fred Williams [mailto:[EMAIL PROTECTED]
Sent: 28 January 2007 18:49
To:
OK, thanks, size of the query shouldn't be a problem then.
Performance is fine as well. This particular query ran in 1.4 secs, which
in my app is fine as it is reporting software and results are expected to
take a bit of time.
This query is the final query of a number of queries that transpose a
Wow! Talk about obfuscated code! I didn't even try to dig deeper than
a quick scan, but could this abomination be broken into multiple update
queries? On the surface it looks like each "group" is unique. If so,
wouldn't a transaction with multiple update statements be much more
efficient and a
"RB Smissaert" <[EMAIL PROTECTED]> wrote:
> Is there any limit on the size of the SQL statements in SQLite?
32-bit integers are used to count things in various places. I
don't really consider that a limit, but some people do.
See http://www.sqlite.org/cvstrac/tktview?tn=2125
There may be other
Is there any limit on the size of the SQL statements in SQLite?
Didn't think this would come into play, but have now come across this
query and wonder if this needs considering:
UPDATE A3Test115_J SET ENTRY_ID_E2 = (SELECT ENTRY_ID FROM GROUP_2 T WHERE
PATIENT_ID = T.PID), READ_CODE_E2 = (SELECT
[EMAIL PROTECTED] wrote:
So I could just use
SELECT ..., SUM(`downloadCount`) FROM ... WHERE `pid` = (SELECT `id`
FROM ...) GROUP
BY `price`;
Is that going to do the same as
SELECT SUM(`downloadCount` FROM (SELECT ... FROM ... WHERE `pid` =
(SELECT ...)) GROUP BY `price`;
?
Yes, these two will
On Saturday, January 27, 2007, 10:10:27 PM, Igor
Tandetnik wrote:
> [EMAIL PROTECTED] wrote:
>> Actually, my query is something like
>> SELECT ... FROM ... WHERE `pid` = (SELECT `id` FROM ...);
>> if i put that group by... will it group all rows, or only those with
>> the
15 matches
Mail list logo