top of page

Workflow Automation Templates

A library of ready-to-use workflow templates to accelerate your data journey

Time Series FE PySpark

Generate time-based features for analysis

Data-cleaning.jpg
Overview

This workflow processes transactional data to create time-series features using PySpark. It enriches datasets with temporal and behavioral metrics, helping analyze user patterns, seasonality, and transaction trends over time.

Details

The workflow begins by loading the dataset containing user transactions. The Time Series Features node computes time-based attributes such as days since the last transaction, days until the next transaction, transaction frequency, and rolling averages.

It also derives features like transaction hour, day of week, week of year, and seasonal indicators, providing a comprehensive temporal view of user activity. The resulting dataset, combining original and newly engineered time-series columns, is displayed using the Print N Rows node.

This workflow enhances analytical depth by enabling models to capture time-driven behaviors and periodic patterns effectively.

bottom of page