Skip to contents

Improved version of aggregate

Source: R/aggreg.R
aggreg.Rd

Improve on miserable aggregate function:

  • ignore NAs

  • return NA if all in group are Inf or NaN

  • no warnings

  • sort by grouping variable and optionally return only result column

  • optionally drop groups with too many NAs

Usage

aggreg(x, by, FUN, drop_by = TRUE, nomiss = NULL)

Arguments

x

Vector to aggregate

by

Vector(s) to group by (if only 1 grouping variable, this doesn't have to be a list)

FUN

Function to summarize with

drop_by

If TRUE, drop the grouping variable and just return a vector; otherwise, return a data frame

nomiss

If not NULL, this represents a proportion of the data in a group that must be non-missing; if this threshold isn't met, the result for the group will be NA

Value

Vector of aggregated values (if drop_by = TRUE), or data frame of groups and aggregated values (if drop_by = FALSE)