Comments on: Business Data Analysis : Comparing Data Populations with SQL http://technology-advisors.net/06/business-data-analysis-data-sql/ Guides for Risk Mgmt, Managed Services, Business Continuity Security, Information Assurance, Enterprise Content Management Tue, 15 Dec 2009 22:18:24 +0000 hourly 1 http://wordpress.org/?v=3.3.2 By: Technology http://technology-advisors.net/06/business-data-analysis-data-sql/comment-page-1/#comment-134 Technology Sun, 14 Jun 2009 05:23:40 +0000 http://technology-advisors.net/?p=109#comment-134 Perfect question! IMO from a technical or IT perspective your query is perfect and the most efficient way to write it and would be the best way if we were dealing with perfect data set. However, from a business analysis perspective, more often than not you will be more efficient with the sub queries especially when you are first <strong>researching </strong>an unknown data population. Here is an example, in the 'Left Join Query' above where there are 30 loans in the in the original population and not in the returned population. Leaving the 'select * from original_loan_pop' allows you to see the details of the loans or data set and easily add or remove fields from the query. This allows flexibility in researching the issue or variance. Perfect question! IMO from a technical or IT perspective your query is perfect and the most efficient way to write it and would be the best way if we were dealing with perfect data set. However, from a business analysis perspective, more often than not you will be more efficient with the sub queries especially when you are first researching an unknown data population.
Here is an example, in the ‘Left Join Query’ above where there are 30 loans in the in the original population and not in the returned population. Leaving the ‘select * from original_loan_pop’ allows you to see the details of the loans or data set and easily add or remove fields from the query. This allows flexibility in researching the issue or variance.

]]> By: Nick Fotopoulos http://technology-advisors.net/06/business-data-analysis-data-sql/comment-page-1/#comment-133 Nick Fotopoulos Sun, 14 Jun 2009 03:46:32 +0000 http://technology-advisors.net/?p=109#comment-133 Why write the select list as sub queries when you don't have any additional joins and where clauses in them. Select * from (select * from original_loan_pop)a join (select * from returned_loan_pop)b on a.loan_num = b.loan_num is the same as Select * from original_loan_pop a join returned_loan_pop b on a.loan_num = b.loan_num ??? Why write the select list as sub queries when you don’t have any additional joins and where clauses in them.

Select * from
(select * from original_loan_pop)a
join
(select * from returned_loan_pop)b
on a.loan_num = b.loan_num

is the same as

Select * from
original_loan_pop a
join
returned_loan_pop b
on a.loan_num = b.loan_num

???

]]>