SQL Tree

I am working with T-SQL and read that CTE is better than subqeries when working with Trees.  I am not familiar with the term Tree with regards to T-SQL.  Any ideas on what this is and how it is used in T-SQL ? Thanks for any feedback.
Who is Participating?
Surendra NathTechnology LeadCommented:
Hi There

To simplify the things over here, the SQL Tree is nothing different from a family tree, that we all are most familiar with.... A parent will have children.. grand children.. grand grand childrens......

in the a tree the parent is called the parent node and the young generation of the family is called the leaf node.

Now, how this tree can be applied in T-SQL...
Let us assume you have a table named schema as below


now the data in the table can be as below

Employee_ID    Employee_Name    Reporting_manager
1                           FRED                          none
2                           Surendra                   FRED
3                           Neo                             FRED
4                           Jarvis                          Surendra

now, if we want to know the reporting hierarchy for the employees... then this is how it will be

Surendra  & Neo will be directly reporting to FRED, and Jarvis will be reporting to Surendra who inturn reports to Fred.

This is a organizational tree, if we need to find out who reports to who kind of quries, it can be done using recursion using CTE or usual subqueries or Cursors...
But the fast apporach is with CTE.
the term 'tree' is sometimes used as a way to describe a hierarchical query, in TSQL these typically require "recursion" and CTEs are a good method for this.

refer to this article perhaps:
a CTE approach is discsssed there (referring to a BOM which is a hierarchy or 'tree')
Racim BOUDJAKDJIDatabase Architect - Dba - Data ScientistCommented:
You can use the following methodologies to represent hierarchies:

> Nested Set
> Materialized Path
> Adjacency

None of the above require recursion.
Keep up with what's happening at Experts Exchange!

Sign up to receive Decoded, a new monthly digest with product updates, feature release info, continuing education opportunities, and more.

Racim BOUDJAKDJIDatabase Architect - Dba - Data ScientistCommented:
If your tree has a high number of levels, recursion is to be avoided to prevent excessive locking.
opinions differ (on almost all sql topics :), e.g.
... the simplicity of adjacency list implementation and management, make adjacency list a preferred way to store hierarchical data in SQL Server versions that support recursive CTEs, that is versions from 2005 onwards.
(emphasis added), see http://explainextended.com/2009/09/25/adjacency-list-vs-nested-sets-sql-server/

the article given earlier also presents alternative approaches.
Racim BOUDJAKDJIDatabase Architect - Dba - Data ScientistCommented:
<<opinions differ (on almost all sql topics :), >>
Fair enough but looking at the example provided, the OP only shows he does not know how to implement nested sets on SQL Server and blames poor performance not realizing he does not understand how to implement them:

> First, he does not understand that right/left are to be on the same table for browsing a specific hierarchy.  He uses an example of a JOIN operation between 2 hierarchies to conclude for poor performance.  Typically, the left and right sides on nested set are columns not tables.
> Second as a consequence of the above, he uses a JOIN operation to justify poor performance
> Third, on a nested set representation, leafs and ancestors are obtained  according the difference between right and left values: all descendants are obtained according to the fact that the difference between the rich and left is always 1.  All ancestors are obtained as having the largest differences between right and left values AND including a specific value.  In other words, the OP says he is using nested set when he is not.

For what it is worth, I have above 10 level deep data warehouses running BI hierarchies returning real time grouped aggregates to report users and the slowest query I have is one second.  This is simply impossible using adjacency lists.

I recommend nested sets when browsing, aggregation and tree depth become important and adjacency lists in other cases because they are easier to implement.

Hope this helps.
Racim BOUDJAKDJIDatabase Architect - Dba - Data ScientistCommented:
fjkaykr11Author Commented:
Great responses.   @Racimo and @NeoJarvis -- thank you for providing specific examples of what I am looking to grasp.
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.