CODEWITHSUNDEEP
r
nir020
nir020

Reputation: 97

Finding and replacing multiple items in R

As a newbie to R i am having to write all my find and replacement statements one line at time (see code below) Is it possible to do this in a more succinct way (i.e. one line only)

YP$gender <- replace(as.character(YP$gender), YP$gender == "Female", "F")   
YP$gender <- replace(as.character(YP$gender), YP$gender == "Male", "M")

Upvotes: 3

Views: 108

Answers (4)

zx8754
zx8754

Reputation: 56219

Looks like you have factor column, so we just need to change the labels, something like this:

YP$gender <- factor(YP$gender, labels = c("F", "M"))

Reproducible example:

x <- factor(c("Female", "Male", "Female"))
x
# [1] Female Male   Female
# Levels: Female Male

#Check the levels
levels(x)
# [1] "Female" "Male"  

# relabel
x <- factor(x, labels = c("F", "M"))
# [1] F M F
# Levels: F M
levels(x)
# [1] "F" "M"

Upvotes: 1

Pedro Henrique S
Pedro Henrique S

Reputation: 115

You can use '%in%' for multiples comparations instead of '=='.

replace(as.character(YP$gender), YP$gender %in% c("Male","Female"), c("M", "F"))

EDIT: Sorry this code won't work as I tought.

But you can use loops to solve it.

YP = c("a","b","b","a","c")

keys = c("a", "b", "c")
rep_value = c("A", "B", "C")

for(index in 1:length(keys) ) {
  sub_key = keys[index]
  sub_rep_value = rep_value[index]

  value_index = which(YP %in% sub_key)
  YP[value_index] = sub_rep_value
}

Upvotes: 1

DeduciveR
DeduciveR

Reputation: 1702

Depends how many targets and replacements you have. If you have a lot then the easiest way is probably to create a merge file with two columns, one with the target and one with the replacement with as many rows as unique elements. If that merge file is called df then the code would look something like:

library(dplyr)
YP <- YP %>%
  merge(df, by = "gender", all.x = T)

If there aren't too many unique instances to replace then instead of using nested ifelse statements, you could use case_when from dplyr. You can chain together the logic using pipes %>%

library(dplyr)
YP %>%
  mutate(gender = case_when(
    gender == "Female" ~ "F",
    gender == "Male"   ~ "M,
    TRUE               ~ gender
  ))

Upvotes: 1

Esben Eickhardt
Esben Eickhardt

Reputation: 3862

If just two replacements use 'ifelse':

YP$gender <- ifelse(as.character(YP$gender) == "Female", "F", "M")

Else I would use left_join:

# Data
df <- data.frame(value = sample(1:3,10, replace = TRUE),
                 gender = sample(c("male", "female", "x"), 10, prob = c(0.4,0.4,0.2), replace = TRUE))

# Creating replacements
replace <- data.frame(gender = c("male", "female"), gender_short = c("m", "f"))

# Making replacements
library(dplyr)
df <- left_join(df, replace)
df

       value gender gender_short
1      1 female            f
2      2 female            f
3      3      x         <NA>
4      2   male            m
5      3 female            f
6      3      x         <NA>
7      3 female            f
8      1      x         <NA>
9      3   male            m
10     3   male            m

Upvotes: 2

Related Questions