tidyselect

tidyselect is an R package that provides a concise, consistent language for selecting columns from data frames and similar objects. It is a core part of the tidyverse, used by dplyr, tibble, and related packages to identify variables in a predictable and readable way. tidyselect does not perform data operations itself; instead, it supplies the mechanism by which column names or positions are chosen for subsequent actions.

The package implements a set of selector helpers and a small domain-specific language. Common selectors include starts_with("prefix"), ends_with("suffix"), contains("text"), and matches("regex"), which select columns by name patterns. Others support numeric or positional selection, such as everything() for all columns, last_col() for the final column, and num_range("x", 1:3) for a numeric sequence of names. There are also helpers for explicit inclusion or exclusion, such as one_of, all_of, and any_of, with all_of requiring all provided names to exist and any_of selecting only those that exist. The where(predicate) selector allows choosing columns that satisfy a given function, such as is.numeric.

tidyselect operates by evaluating selectors against a data frame’s column names and returns the positions of matching columns, which dplyr then uses in its verbs. It supports non-syntactic names through backticks and works with standard data frames as well as tibbles. While intended for column selection, tidyselect emphasizes ease of use, readability, and compatibility across the tidyverse, rather than exhaustive data manipulation capabilities.