http://www.sqlpower.ca/forum/forums/show/204.page JForum - http://www.jforum.net (
select mobid,
max(case when seq=1 then merchant else null end) as min_merchant,
max(case when seq=1 then price else null end) as min_price,
max(case when bseq=1 then merchant else null end) as max_merchant,
max(case when bseq=1 then price else null end) as max_price
from
(
select row_number() over (partition by mobid order by price) as seq,row_number() over (partition by mobid order by price desc) as bseq,mobid, merchant,price
from tbl_merchant
)t
where (seq=1 or bseq=1) and mobid in(select MOBID from tbl_product where Client_Name='wilson')


group by mobid
) as t1 INNER JOIN tbl_product as t2 on t2.mobid=t1.mobid

In the above query i need to add my_Price field where
my_price is where my name matches for that product




]]> http://www.sqlpower.ca/forum/posts/preList/2691/6059.page http://www.sqlpower.ca/forum/posts/preList/2691/6059.page GMT
for the deduping it could be interesting to have to option to delete a record because it is not needed. It might not always be a match or a unique record but just a useless record.

Benjamin]]> http://www.sqlpower.ca/forum/posts/preList/2546/5687.page http://www.sqlpower.ca/forum/posts/preList/2546/5687.page GMT
I often have this problem: I many cases, I run a match where one row matches with hundreds others rows. The results is that is very very long to draw results.
One suggestion could be: if a match includes too many rows (more than 30 for example) only the 30 firsts are printed. Then it is possible to see which element causes this large amount of matches and update the transformation to solve it.

I think that in this case, it's not really useful to have hundred times the same row...

Just another comment, I always have to select two times the database name, table name ... in check lists when creating a new project or repository... The first time you select it, nothing appear in the box.

(I've got the 9.5 version)

Simon
]]> http://www.sqlpower.ca/forum/posts/preList/2427/5306.page http://www.sqlpower.ca/forum/posts/preList/2427/5306.page GMT java.lang.runtimeexception y no me deja usarla nunca , le agradecere su ayuda]]> http://www.sqlpower.ca/forum/posts/preList/2425/5299.page http://www.sqlpower.ca/forum/posts/preList/2425/5299.page GMT
have you thought about creating a Kettle plugin for DQGuru? This could help to do data cleansing during ETL process. Users could set up the processes in DQGuru and then call them out of Kettle.

Benjamin]]> http://www.sqlpower.ca/forum/posts/preList/2364/5017.page http://www.sqlpower.ca/forum/posts/preList/2364/5017.page GMT The user should be able to set how many words to match on.
The word could appear in any position within the column, for the match to occur.

It would be similar to using SQL to search a column with a WHERE clause similar to below
where column_name like '%word%'

]]> http://www.sqlpower.ca/forum/posts/preList/2352/4987.page http://www.sqlpower.ca/forum/posts/preList/2352/4987.page GMT
here are my suggestions & ideas for DQGuru as discussed yesterday. They are all filed in Bugzilla already:

- Enable CSV writer to accept all data formats as input (Bug 1798)
[quote]Right now CSV writer only takes strings as inputs, all other formats need a
extra step to convert to string first. It would be nice if CSV writer would
accept all formats as input.[/quote]

- Save layout of transformation
[quote]It would be nice if the layout of a transformation would get saved. Right now the design of a transformation doesn't get saved after closing and opening it.
You have to unfold all steps manually again.[/quote]]]> http://www.sqlpower.ca/forum/posts/preList/2341/4959.page http://www.sqlpower.ca/forum/posts/preList/2341/4959.page GMT
Is there an easy way for the community to create new transformation functions and to share them? Ideally, one would be able to go to a community area, review the available functions, and import them into the DQ Guru as needed.

Thanks]]> http://www.sqlpower.ca/forum/posts/preList/2308/4836.page http://www.sqlpower.ca/forum/posts/preList/2308/4836.page GMT
I've been playing with DQ Guru. As a non-technical individual, it would be nice to be able to create 'super transformations' that are made up of the base transformations already contained within DQ Guru. I would then be able to reuse some of the more complex cleansing processes that I've developed.

Thanks,]]> http://www.sqlpower.ca/forum/posts/preList/2307/4835.page http://www.sqlpower.ca/forum/posts/preList/2307/4835.page GMT
Consider including a REFRESH button to refresh the meta data cache or not caching at all.

Best wishes.

Donn]]> http://www.sqlpower.ca/forum/posts/preList/2212/4493.page http://www.sqlpower.ca/forum/posts/preList/2212/4493.page GMT
When working in database schemas with large number of tables, this makes it difficult to find the target table.

I have not been bothered by this problem in other dialog locations, but it is probably a good standard practice.

Thanks.

Donn]]> http://www.sqlpower.ca/forum/posts/preList/2211/4492.page http://www.sqlpower.ca/forum/posts/preList/2211/4492.page GMT http://www.sqlpower.ca/forum/posts/preList/2055/3903.page http://www.sqlpower.ca/forum/posts/preList/2055/3903.page GMT http://www.sqlpower.ca/forum/posts/preList/2054/3902.page http://www.sqlpower.ca/forum/posts/preList/2054/3902.page GMT
It would be nice to be able to sort the graphs based on whatever values were chosen to be displayed in the graphs.

For example, after i have chosen to display the 'first_name' field of each record, I thought it would be nice to be able to sort the graphs by first name so that I can search for a particular Match Group better.

Would that be easy to implement?

Thanks.]]> http://www.sqlpower.ca/forum/posts/preList/2052/3900.page http://www.sqlpower.ca/forum/posts/preList/2052/3900.page GMT Version: Nightly build 23072008

When writing the matches to the match pool is there a way that the new matches can be added and not receive an error msg for the matches that are already there?

Currently if you run the match engine once, make some modifications to the munge processes and run the match engine again, you will receive an error if the same match was found.

I think the engine should be able to skip the record already in the match pool and move on to load the new matches found without getting an error message. ]]> http://www.sqlpower.ca/forum/posts/preList/2029/3212.page http://www.sqlpower.ca/forum/posts/preList/2029/3212.page GMT