Wednesday, March 31 2021
4:00pm-5:00pm Eastern Time (ET)
Title: dplyr: one language, many implementations
Abstract: One of dplyr's lesser known features is that it works with data stored in a wide range of ways, translating dplyr verbs into a variety of other computational frameworks. In his talk, Hadley will talk about three important backends: dtplyr, dbplyr, and multidplyr. These allow dplyr to seamlessly scale up to handle every larger dataset:
- dtplyr uses the fantastic data.table package to quickly work with large in-memory datasets;
- dbplyr converts your R code to SQL so you can work with data of any size in a relational database; and
- multidplyr allows you to easily take advantage of every core on your computer.
Hadley also will discuss recent community contributions that extend these backends to key tidyr verbs, and share why he thinks the idea of separation description from computation is such a powerful idea.