impute_dt: Impute missing values with mean, median or mode

In tidyfst: Tidy Verbs for Fast Data Manipulation

Impute missing values with mean, median or mode

Description

Impute the columns of data.frame with its mean, median or mode.

Usage

impute_dt(.data, ..., .func = "mode")

Arguments

.data	A data.frame
...	Columns to select
.func	Character, "mode" (default), "mean" or "median". Could also define it by oneself.

Value

A data.table

Examples

Pclass <- c(3, 1, 3, 1, 3, 2, 2, 3, NA, NA)
Sex <- c('male', 'male', 'female', 'female', 'female',
         'female', NA, 'male', 'female', NA)
Age <- c(22, 38, 26, 35, NA,
         45, 25, 39, 28, 40)
SibSp <- c(0, 1, 3, 1, 2, 3, 2, 2, NA, 0)
Fare <- c(7.25, 71.3, 7.92, NA, 8.05, 8.46, 51.9, 60, 32, 15)
Embarked <- c('S', NA, 'S', 'Q', 'Q', 'S', 'C', 'S', 'C', 'S')
data <- data.frame('Pclass' = Pclass,
 'Sex' = Sex, 'Age' = Age, 'SibSp' = SibSp,
 'Fare' = Fare, 'Embarked' = Embarked)

data
data %>% impute_dt() # defalut uses "mode" as `.func`
data %>% impute_dt(is.numeric,.func = "mean")
data %>% impute_dt(is.numeric,.func = "median")

my_fun = function(x){
  x[is.na(x)] = (max(x,na.rm = TRUE) - min(x,na.rm = TRUE))/2
  x
}
data %>% impute_dt(is.numeric,.func = my_fun)

tidyfst documentation built on Sept. 16, 2024, 9:06 a.m.