Course description: IBM Netezza Analytics for Developers

Netezza allows you to extend SQL by using user-defined extensions (UDXs), as well as User-Defined Analytic Processes (UDAPs). UDXs can be thought of as the user-defined counterparts of built-in SQL Functions. They are called from SQL in the same manner and follow the same guidelines for input and output. UDAPs, although called from SQL similarly to UDXs, are actually applications that run when called. The UDAP concept allows a Netezza developer to implement a freestanding, executable data-processing program, that runs "out of process", that is, outside the system, and register it in a database.

This course will teach you how to develop User Defined Extensions including: User Defined Functions (UDFs), User Defined Aggregates (UDAs), User Defined Table Functions (UDTFs), and User Defined Analytic Processes (UDAPs). You will develop these User Defined Extensions using the Netezza command line utilities to compile and register these in-database analytics.

View this course in other countries

Roadmaps that contain this course are:

• PureData System for Analytics, powered by Netezza
• Netezza Appliance

Back to top

This advanced course is intended for developers and programmers who want to embed in-database analytics on Netezza.

Back to top

You should have the following background:

• Working knowledge of UNIX or Linux
• Working knowledge of Data Warehousing concepts
• Knowledge of C or C++
• Ability to use the VI Editor
• Knowledge of Java
• Familiarity with Eclipse

Back to top

• Write a user defined function (UDF) in C++ to extend the capabilities of SQL
• Write a user defined aggregate (UDA) in C++ to implement the various phases of aggregate evaluation, such as initialization, accumulation, and merging
• Write a user defined table function (UDTF) in C++ enabling you to process one/many rows to return a table shape composed of many rows/columns
• Manage user defined functions, aggregates and table functions and shared libraries (such as granting permissions)
• Write a user defined analytic process (UDAP) in Java to extend the capabilities of SQL and run an analytic out-of-process. Additionally, be aware that UDAPs can be developed in other programmatic languages
• Know the features of the Netezza Plug-in for Eclipse

Back to top

Agenda

Day 1

• Unit 1: Introduction to IBM Netezza Analytics
• Unit 2: Establishing the Programming Environment
• Exercise 1: Connecting to the training system
• Unit 3: Writing a simple UDF
• Exercise 2: Build the Hello World UDF
• Unit 4: Working with data types
• Exercise 3: Implement date and time helper functions
• Unit 5: Accessing UDFs via cross database access and stored procedures
• Exercise 4: UDFs from stored procedures

Day 2

• Unit 6: Advanced topics in user-defined functions
• Unit 7: User-defined table functions
• Exercise 5: Implement the unpivot function
• Unit 8: Working with user-defined aggregates
• Exercise 6: Implement the user-defined aggregates
• Unit 9: User-defined analytical processes
• Unit 10: UDAP using scalar functions (UDFs)
• Exercise 7: Build Hello World using UDAP in Java
• Unit 11: UDAP using table functions (UDTFs)
• Exercise 8: Build the unpivot function as a UDAP

Day 3

• Unit 12: UDAP using aggregate (UDAs)
• Exercise 9: Implement the common aggregate as a UDAP
• Unit 13: Debugging user-defined functions and aggregates
• Exercise 10: Modify an existing UDF to include error checking
• Unit 14: How to administer the system
• Exercise 11: Back up your analytics database

Back to top

Practical Work:

This course is structured in a lecture/lab format. The hands-on sessions form a vital and integral part of the course.

March 2012

V1.1

Back to top