Skip to contents

Compute Fast Summary Statistics by Group

compute_fastsummary.Rd

`compute_fastsummary()` computes summary statistics for selected columns of a dataset, optionally grouped by one or more variables. It allows the user to specify a set of functions to apply, either from a predefined set or custom formulas/functions.

Usage

compute_fastsummary(
  data,
  cols,
  fns = NULL,
  groups,
  output = c("long", "wide"),
  tbl = FALSE
)

Arguments

data

A `data.table`, `data.frame`, or tibble. The dataset on which to compute the summaries. If not a `data.table`, it will be converted internally for computation. The result will be returned in the same class as the input (unless `tbl = TRUE`).

cols

A character vector. Names of the columns to summarize.

fns

Optional. Either:

  • `NULL` (default): use all default functions defined by `define_fns()`.

  • A character vector of function names matching `define_fns()`.

  • A list of functions or formulas, possibly mixed with character names referring to `define_fns()`.

groups

A character vector. Column(s) by which to group the data before computing the summary statistics.

output

Character. Either `"long"` (default) or `"wide"` to specify the output format. `"long"` returns one row per group per summary statistic, `"wide"` returns one row per group with multiple columns for each summary statistic.

tbl

Logical. If `TRUE`, converts the result to a tibble (`tibble::as_tibble()`).

Value

A dataset containing the summary statistics for the selected columns. The output will be either long or wide depending on the `output` argument. The returned object will match the class of the input `data` (unless `tbl = TRUE`).

Details

The function constructs the summary calls efficiently using `bquote()` and evaluates them within the `data.table` environment. This allows for fast computation even with large datasets. Custom functions can be supplied as formulas (e.g., `~ mean(.x, na.rm = TRUE)`) or as pre-defined function names from `define_fns()`.

Examples

if (FALSE) { # \dontrun{
library(data.table)
dt <- data.table(x = rnorm(100), y = rnorm(100), group = sample(1:2, 100, TRUE))
# Compute mean and sd by group
compute_fastsummary(dt, cols = c("x", "y"), fns = c("mean", "sd"), groups = "group")

# Use a custom function
compute_fastsummary(
  dt,
  cols = "x",
  fns = list(mean = ~mean(.x, na.rm = TRUE)),
  groups = "group",
  output = "long",
  tbl = TRUE
)
} # }