Re: [sqlite] Student's t-test table

[Giuseppe Costanzi]

Thanks Simon, I know how to perform the importation.
However thanks of the idea.
I start looking for it.

On Sun, Nov 16, 2014 at 8:46 AM, Simon Slavin  wrote:
>
> On 16 Nov 2014, at 7:39am, Giuseppe Costanzi  
> wrote:
>
>> do you know if a database exists in sqlite with the values, degrees of
>> fredom vs  probability (alfa) of the t Student?
>
> If you can find one in text or .csv format, we can tell you how to import it.
>
> Simon.
> ___
> sqlite-users mailing list
> sqlite-users@sqlite.org
> http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
___
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users

Re: [sqlite] Student's t-test table

[Simon Slavin]

On 16 Nov 2014, at 7:39am, Giuseppe Costanzi  wrote:

> do you know if a database exists in sqlite with the values, degrees of
> fredom vs  probability (alfa) of the t Student?

If you can find one in text or .csv format, we can tell you how to import it.

Simon.
___
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users

[sqlite] Student's t-test table

[Giuseppe Costanzi]

Hi to everybody,

do you know if a database exists in sqlite with the values, degrees of
fredom vs  probability (alfa) of the t Student?

I need it to calculate confidence intervals to compare two means

regards

bc1966
___
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users

Re: [sqlite] Discrepancy with indexing and WHERE clause with AND/OR

[RSmith]

On 2014/11/14 16:32, RP McMurphy wrote:


On Tue, 11/11/14, RP McMurphy  wrote:

  > If you can provide
  any examples where ANALYZE makes a query slower, I
  suspect the developer team would like
  > to see them.
  >
  
  After we run analyze and then

  let the process run for a while the DB
  contents change - and it can change quite
  considerably depending
  upon what is
  happening. I suspect that the analyze data gets stale, but
  I
  don't know how to track such things in
  sqlite. Anyhow we can't keep running
  analyze every few minutes because it takes a
  long time to run with our DB
  and it appears
  to block all other actions until it is done.


Firstly, that is a considerably different problem to your initial claim that ANALYZE makes queries slower. That said, it is an 
understandable concern.  When you say the data changes, does the shape actually change? Because all that matters to analyze is the 
shape of the data, not the data itself.


As an example, consider the following three small tables:

   A:
   ID, No, Name,   Value, XT,  AX
   1,  1,  John,   34,Panthera Leo,2014-11-12 09:44:26
   2,  2,  Chris,  88,Panthera Pardis, 2014-11-13 04:17:43
   3,  3,  Martin, 20, etc...

   B:
   ID, No, Name,  Value, XT,   AX
   1,  1,  John,  34,,
   2,  X1, John,  ,  Panthera Leo,
   3,  A1, John,  ,  , 2014-11-12 09:44:26
   4,  2,  Chris, 88,,
   5,  X2, Chris, ,  Panthera Pardis,
   6,  A2, Chris, ,  , 2014-11-13 04:17:43
   etc.

   C:
   ID,  No, Name, Value, XT,  AX
   106, 62, Andrew,   18,Loxodonta Africana,  -00-00 00:00:00
   107, 677,James,12,Podocarpus Falcatus, 2014-11-09 14:27:12
   108, 176612, Flemming, 40,Tectonia Grandis,2014-11-14 03:31:55
   etc.



From table A to B the data did not change significantly, but the shape of it changed significantly in respect of cardinality, value 
types, fill rate etc.  From A to C the data is completely different, but the shape is exactly the same.


Running Analyze for table A which shape-shifts into B over time (which is highly unlikely) will certainly see the values become 
"stale". Running Analyze for A which eventually changes into C (very likely) will have minimal effect on the outcome of the Analyze 
data and hence on the outcome of the Query Planner decisions.


To re-iterate the question posed earlier - If you can show that the QP makes a worse decision after analyzing a table - OR - that 
the decision degrades over time for a table of which the shape does not change effectively - I am sure the case would be considered 
as undesirable and the devs would probably be interested in doing something about it.


Other options include forcing the use of an Index or Indices (see "USING" in 
the SQLite SQL docs) or using cross-joins and the like.



  A this point we are
  considering writing VDBE code directly and bypassing the
  parser. Has anyone else done this? Is it going
  to be a huge ugly
  can-of-worms if we do
  that?


Not just a can of worms - an insane one, and what is worse, you will immediately lose any benefit of periodic updates and all the 
other advances in technology so frequently visited upon SQLite because your version is no longer interchangeable with the new ones. 
Further to that, this speaks of the willingness to take 3 steps in the wrong direction to avoid one step in the right direction - 
maybe some past experience influence or some other unknowns which are not immediately obvious now is influencing this thought? (I 
only ask because this is not uncommon).


If you do have a table of which the shape keeps changing to the point it needs re-analyzing every few minutes, then you are working 
with the worst DB in history and should really make your own native storage solution - but I am 90% certain this is not the case and 
you are simply overlooking something or not doing something efficiently - which we will be glad to help with if you post schemata 
and maybe link an example DB or such.


Unlike me, some other people on here are extremely knowledgeable on SQL and specifically the inner workings of SQLite - Tell us the 
full story, we might save you hundreds of hours.


___
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users

Re: [sqlite] Discrepancy with indexing and WHERE clause with AND/OR

[Simon Slavin]

On 14 Nov 2014, at 2:32pm, RP McMurphy  wrote:

> After we run analyze and then
> let the process run for a while the DB
> contents change - and it can change quite
> considerably depending
> upon what is
> happening
> 
> I suspect that the analyze data gets stale, but
> I don't know how to track such things in
> sqlite. Anyhow we can't keep running
> analyze every few minutes because it takes a
> long time to run with our DB
> and it appears
> to block all other actions until it is done.

Okay.  ANALYZE isn't used for specific information about every piece of data.  
It's used to evaluate the 'chunkiness' of each column and index.  For instance 
one column may only ever have two values in it -- "yes" or "no" -- whereas 
another column may have a different value in it for each row.

So unless the character of your data changes, if you run ANALYZE once when you 
have realistic data, you shouldn't need to run it again even if your tables 
grow to 100 times their original size.  You don't need to run ANALYZE under 
normal operation.  Perhaps once a year in a yearly maintenance run or 
something.  Or if you create a new index so SQLite can see how useful that 
index will be for certain operations.

As others have shown you using EXPLAIN QUERY PLAN, SQLite is itself capable of 
doing the expansion you describe to get an efficient query.  I would definitely 
not recommend you try beating the SQLite internals which have been developed 
and improved for years.  It's going to be more effective to see if you can spot 
some point at which your query starts executing slowly.

Simon.
___
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users

Re: [sqlite] querying with BLOB in WHERE clause, possible?

[Stephan Beal]

On Fri, Nov 14, 2014 at 3:57 PM, bjdodo  wrote:

> I've got it now. E.g. for delete you can do this:
>

A minor fix:

the 'try' needs to be moved to...


>
> int argidx = 0;
> SQLiteStatement statement = mDB.compileStatement(sql);
>

here. Because...


> throw new RuntimeException("unexpected argument type);
>

if that happens then you're not closing it.


> try {
> return statement.executeUpdateDelete();
> } finally {
> statement.close();
> }
>

-- 
- stephan beal
http://wanderinghorse.net/home/stephan/
http://gplus.to/sgbeal
"Freedom is sloppy. But since tyranny's the only guaranteed byproduct of
those who insist on a perfect world, freedom will have to do." -- Bigby Wolf
___
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users

Re: [sqlite] querying with BLOB in WHERE clause, possible?

[bjdodo]

I've got it now. E.g. for delete you can do this:


String sql = "DELETE FROM " + table +
(!TextUtils.isEmpty(whereClause) ? " WHERE " + whereClause :
"");

int argidx = 0;
SQLiteStatement statement = mDB.compileStatement(sql);
if (whereArgs != null) {
for (WhereArg wa : whereArgs) {
Object o = wa.getParamValueRaw();
if (o == null) {
statement.bindNull(++argidx);
} else if (o instanceof String) {
statement.bindString(++argidx, (String) o);
} else if (o instanceof Long) {
statement.bindLong(++argidx, (Long) o);
} else if (o instanceof Double) {
statement.bindDouble(++argidx, (Double) o);
} else if (o instanceof byte[]) {
statement.bindBlob(++argidx, (byte[]) o);
} else {
throw new RuntimeException("unexpected argument type);
}
}
}

try {
return statement.executeUpdateDelete();
} finally {
statement.close();
}







--
View this message in context: 
http://sqlite.1065341.n5.nabble.com/querying-with-BLOB-in-WHERE-clause-possible-tp44550p79220.html
Sent from the SQLite mailing list archive at Nabble.com.
___
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users

Re: [sqlite] Missing SQLite.Designer.SQLiteDataViewSupport?

[echizen]

I have installed the version 1.0.94.0 and I get the same problem, telling me
"Could not find Any Appropriate resources for the culture or the Specified
neutral culture. Make sure" SQLite.Designer.SQLiteDataViewSupport2012.xml
"was embedded or linked into Correctly assembly "SQLite.Designer" at compile
time, or all the satellite assemblies That required are loadable and fully
signed. "

What could it be?



--
View this message in context: 
http://sqlite.1065341.n5.nabble.com/Missing-SQLite-Designer-SQLiteDataViewSupport-tp78087p79225.html
Sent from the SQLite mailing list archive at Nabble.com.
___
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users

Re: [sqlite] Discrepancy with indexing and WHERE clause with AND/OR

[RP McMurphy]

I am resending this message below from 3 days ago because it never made it to 
the list.

RP

PS: Messages seem to take a long time to go through the gmane system, at least 
half a day and sometimes more in my experience so far.


On Tue, 11/11/14, RP McMurphy  wrote:

 Subject: Re: Discrepancy with indexing and WHERE clause with AND/OR
 To: sqlite-users@sqlite.org
 Date: Tuesday, November 11, 2014, 6:31 PM
 
 > If you can provide
 any examples where ANALYZE makes a query slower, I
 suspect the developer team would like
 > to see them.
 > 
 
 After we run analyze and then
 let the process run for a while the DB
 contents change - and it can change quite
 considerably depending
 upon what is
 happening. I suspect that the analyze data gets stale, but
 I
 don't know how to track such things in
 sqlite. Anyhow we can't keep running
 analyze every few minutes because it takes a
 long time to run with our DB
 and it appears
 to block all other actions until it is done.
 
 A this point we are
 considering writing VDBE code directly and bypassing the
 parser. Has anyone else done this? Is it going
 to be a huge ugly
 can-of-worms if we do
 that?
 
 RP
 
  
 
 
___
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users