This repository provides materials for a session that is part of the I2DS Tools for Data Science workshop run at the Hertie School, Berlin in November 2022. The student-run workshop is part of the course Introduction to Data Science taught by Simon Munzert at the Hertie School, Berlin, in Fall 2022.
This session will introduce you to tidy temporal data and tidy forecasting with fable. Forecasting based on temporal data is highly important for many areas of policy-making and business and thus a highly relevant skill for a data scientist. While tsibble provides the data structure based on the principles of tidy data, fable is a complementary package that uses tsibble objects and allows us to specify models to create forecasts.
The goals of this session are to (1) equip you with conceptual knowledge about temporal data and the tsibble package, (2) show you a forecasting workflow using fable, and (3) provide you with a hands-on exercise to allow you to apply your tsibble and fable skills.
For those who are keen to learn more, please check out the list of further resources.
- tsibble overview at dplyr.tidyverse.org
- fable overview at dplyr.tidyverse.org
- Cran Webpage by Tsibble Creator
- Cran Package for tsibble
- Cran Package for fable
- Hands-on fable tutorial by Package maintainer
- Forecasting Principles and Practice (Online Textbook)
The material in this repository is made available under the MIT license.
In general, both authors collaborated closely together. The division below was made mostly for the efficient execution of the concept that was developed together.
Justus v. Samson-Himmelstjerna prepared the presentation, including slides, and post-processing of the recording.
Oskar Krafft prepared the practice materials and the code for the slides.