In MilesMcBain/wisegroup: Safe grouping wrappers for group-aware tidy functions

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "100%"
)

wisegroup

Now I am become death, destroyer of groups....

Never be burned by forgetting to ungroup() again.

This packages contains alternative versions of common group-aware tidyverse functions that either:

automatically ungroup the returned data OR
allow the developer to make explicit that groups are being carried forward, signalling the potential for group related errors with a '...' variant of the orignal function, e.g. mutate..., summarise... etc.

Installation

remotes::install_github("milesmcbain/wisegroup")

Usage

wisegroup is designed to mask tidyverse functions in your global environment. It is recommended that you use this package in conjunction with conflicted. This means that the order of the library calls will not matter, and that the masking is made explicit like so:

library(conflicted)
library(wisegroup)
library(tidyverse)
library(nycflights13)

conflict_prefer("summarise", "wisegroup")
conflict_prefer("mutate", "wisegroup")

Just be lazy. conflicted will let you know when you need to disambiguate.

So now you have automatically dropped groups after summarising over multiple grouping columns:

grouped_no_more <-
  flights %>%
  group_by(year, month, day) %>%
  summarise(max_delay = max(dep_delay, na.rm = TRUE))

grouped_no_more

You indicate the grouping is sticking around with the ... suffix

still_grouped <-
  flights %>%
  group_by(year, month, day) %>%
  summarise...(max_delay = max(dep_delay, na.rm = TRUE))

still_grouped

and with mutate:

summarised_over_mutates_groups <-
  flights %>%
  group_by(year, month, day) %>%
  mutate...(day_dep_variance = var(dep_delay, na.rm = TRUE)) %>%
  summarise(max_var = max(day_dep_variance))

summarised_over_mutates_groups