Conference Schedule - PGCon 2023

Back

Data science with Postgres - introductory training

Date: 2023-06-02
Time: 13:00–16:00
Room: DMS 1150
Level: Beginner

All your data is on your database, so why do you need to use a third-party programming language to process your data analysis?

Now you can do it in the database! With the TimescaleDB toolkit, you can process your data analysis directly in the database.

TimescaleDB is a time-series extension for Postgresql. It also has a data toolkit as an additional extension, which eases all things analytics when using TimescaleDB, focusing on developer ergonomics and performance.

The training will be a live coding session with the following sections:

  • How TimescaleDB partitions the time-series data to make data processing fast.
  • The toolkit extension: An overview of the features and functions that makes data analysis easy with Postgres.
  • How the pipeline operator brings the power of functional programming directly to the SQL language.
  • Simple finance analysis: Creating candlesticks as know as OHLC charts.
  • Correlation matrix: How to create a correlation matrix in Postgresql from the correlation coefficient.
  • Downsampling: Using the LTTB function to reduce large datasets without losing the visual similarity with the original data.

The training will be exploring an open-weather dataset for practical analysis. Allowing the participants to learn by doing and assimilating the usage of the functions from examples.

There is no video for this session.

Speaker

Jônatas Davi Paganini