gsum {timeplyr}

Grouped statistical functions.

Description

These functions are wrappers around the collapse equivalents but always return a vector the same length and same order as x.
They all accept group IDs for grouped calculations.

Usage

gsum(x, g = NULL, na.rm = TRUE, ...)

gmean(x, g = NULL, na.rm = TRUE, ...)

gmin(x, g = NULL, na.rm = TRUE, ...)

gmax(x, g = NULL, na.rm = TRUE, ...)

gsd(x, g = NULL, na.rm = TRUE, ...)

gvar(x, g = NULL, na.rm = TRUE, ...)

gmode(x, g = NULL, na.rm = TRUE, ...)

gmedian(x, g = NULL, na.rm = TRUE, ...)

gfirst(x, g = NULL, na.rm = TRUE, ...)

glast(x, g = NULL, na.rm = TRUE, ...)

gnobs(x, g = NULL, ...)

Arguments

x

An atomic vector.

g

Group IDs passed directly to collapse::GRP(). This can be a vector, list or data frame.

na.rm

Should NA values be removed? Default is TRUE.

...

Additional parameters passed on to the collapse package equivalents, fsum(), fmean(), fmin(), fmax(), fsd(), fvar(), fmode(), fmedian(), ffirst(), flast() and fnobs()

Value

A vector the same length as x.

Examples

library(timeplyr)
library(dplyr)
library(ggplot2)

# Dplyr
iris %>%
  mutate(mean = mean(Sepal.Length), .by = Species)
# Timeplyr
iris %>%
  mutate(mean = gmean(Sepal.Length, g = Species))

# One can utilise pick() to specify multiple groups
mpg %>%
  mutate(mean = gmean(displ, g = pick(manufacturer, model)))

# Alternatively you can create a unique ID for each group
mpg %>%
  add_group_id(manufacturer, model) %>%
  mutate(mean = gmean(displ, g = group_id))

# Another example

iris %>%
  add_group_id(Species, .name = "g") %>%
  mutate(min = gmin(Sepal.Length, g = g),
         max = gmax(Sepal.Length, g = g),
         sum = gsum(Sepal.Length, g = g),
         mean = gmean(Sepal.Length, g = g)) %>%
  # The below is equivalent to above
  mutate(min2 = min(Sepal.Length),
         max2 = max(Sepal.Length),
         sum2 = sum(Sepal.Length),
         mean2 = mean(Sepal.Length),
         .by = Species) %>%
  distinct(Species,
           min, min2,
           max, max2,
           sum, sum2,
           mean, mean2)