Getting started

In basetable: Fast and Memory-Efficient Base R Table Manipulation

library(basetable)

basetable provides a base-style front end for tabular work with a native C++ table engine. The core design goals are explicit semantics, compact naming, and equal support for nested calls and native |> pipelines.

The canonical compact core

Most workflows need only a small set of regular verbs:

| Task | Canonical functions | |:--|:--| | Inspect | preview(), describe(), missingness() | | Keep rows | subset() | | Keep or remove columns | pick(), drop() | | Create columns | transform() | | Order rows | orderrows() | | Summarise | aggregate(), summaries(), count() | | Join | merge(), antimerge(), semimerge() | | Reshape | tolong(), towide() | | Validate | assertcomplete(), assertunique(), assertrows() |

These are canonical names, not one layer of aliases among several grammars. Each function accepts a table as its first argument, uses character vectors for column specifications, and returns a bt_table. Those regular rules make pipelines compact for people and make function selection predictable for program-generating tools such as language models.

The package is intentionally focused on in-memory table manipulation, exploration, and validation. Read and write files with base R or a dedicated I/O package, then pass the resulting data frame to basetable. Domain-specific parsing and validation remain the responsibility of focused packages.

subset(mtcars, cyl == 6, select = c("mpg", "hp", "wt"))

mtcars |>
  pick(c("mpg", "hp", "wt", "cyl")) |>
  aggregate(by = "cyl", value = c("mpg", "hp"), fun = mean)

Namespaces stay explicit

Compact names intentionally overlap with base R and other table packages. If more than one is attached, qualify the function rather than introducing a new alias or changing grammars mid-pipeline.

mtcars |>
  basetable::subset(cyl == 6) |>
  basetable::pick(c("mpg", "hp", "wt")) |>
  basetable::transform(power = hp / wt)

basetable documentation built on Sept. 18, 2026, 9:13 a.m.