PostgreSQL for business intelligence and big data analysis

This course takes 3 days – 8 hours per day (on site) or 5 hours per day (online).
The only difference is that there will be less practical exercises in an online course.  However, we can share them with you so you can do them on your own. You can ask our consultants for feedback or help, if needed.

DAY 1

Database Design

• Preparing data structures for analysis
• Data structures for data warehousing
• Normalization vs. denormalization
• Handling faulty data

Importing data

• Bulk loading
• LOGGED vs. UNLOGGED tables
• Tuning imports

Aggregating data

• GROUP BY and HAVING
• Optimizing aggregations
• Ordered sets and hypothetical aggregates
• Optimizing the order of aggregation
• GROUP SETS (ROLLUP, CUBE, etc.)

DAY 2

Analytics and windowing functions

• Simple windows (rank, dense_rank, etc.)
• Frame clauses and sliding windows
• Performance tuning

Timeseries analysis

• Storing timeseries data efficiently
• Frame clauses and sliding windows
• Anomaly detection
• Partitioned time series
• Timeseries prediction and forecasting
• Optional: Diving into MobilityDB 
  • Handling and analyzing GPS tracks
  • Storing tracking data in PostgreSQL

DAY 3

Serverside business logic

• Including server side modules
• Data mining and external modules
• Writing your own aggregation functions

Performance tuning

• PostgreSQL for business intelligence
• Detecting and optimizing slow queries
• Optimizing large queries