How to Replace Values in R with Examples

Replacing values in a dataset is a frequent operation in data manipulation and analysis. Whether you are handling missing values, correcting erroneous entries, or transforming data, the R programming language provides various methods to replace values efficiently. In this article, we will explore multiple techniques to replace values in R, each illustrated with practical examples.

1. Using Subsetting and Assignment

A simple way to replace values in a vector or a dataframe is by using subsetting and assignment.

# Example for a Vector
v <- c(1, 2, 3, 4, 5)
v[v == 2] <- 20
print(v)  # Output: 1 20 3 4 5

# Example for a DataFrame
df <- data.frame(
  ID = c(1, 2, 3),
  Value = c(10, 20, 30)
)

df$Value[df$Value == 20] <- 200
print(df)  # Output: ID Value 1 10 2 200 3 30

2. Using the ifelse( ) Function

The ifelse() function is versatile and can be used to replace values based on a condition.

# Example for a Vector
v <- c(1, 2, 3, 4, 5)
v <- ifelse(v == 3, 30, v)
print(v)  # Output: 1 2 30 4 5

# Example for a DataFrame
df$Value <- ifelse(df$Value == 10, 100, df$Value)
print(df)  # Output: ID Value 1 100 2 200 3 30

3. Using the dplyr Package

The dplyr package provides the mutate() and replace() functions to replace values in a dataframe efficiently.

library(dplyr)

# Using mutate and ifelse together
df <- df %>% mutate(Value = ifelse(Value == 30, 300, Value))

# Using replace function
df$Value <- replace(df$Value, df$Value == 100, 1000)

4. Using the na_if( ) Function to Replace with NA

na_if() from dplyr can be used to replace specific values with NA.

# Example
df$Value <- na_if(df$Value, 300)

5. Using the plyr Package’s mapvalues( ) Function

mapvalues() function from plyr package is helpful when you need to replace multiple specific values.

library(plyr)

v <- c(1, 2, 3, 4, 5)
v <- mapvalues(v, from = c(2, 4), to = c(20, 40))

In-Depth Examples:

A. Nested ifelse( ) for Multiple Replacements

v <- c(1, 2, 3, 4, 5)
v <- ifelse(v == 1, 10, ifelse(v == 3, 30, v))

This code replaces 1 with 10 and 3 with 30 in the vector v.

B. Using mutate( ) and case_when( ) Together

df <- df %>% mutate(Value = case_when(
  Value == 1000 ~ 10000,
  Value == 200 ~ 2000,
  TRUE ~ Value
))

This code replaces 1000 with 10000 and 200 with 2000 in the Value column of the dataframe df.

C. Complex Replacement in DataFrames

When dealing with dataframes, sometimes, you need to replace values in multiple columns based on conditions in one or more columns.

df <- df %>% mutate(
  ID = ifelse(ID == 1 & Value == 10000, 10, ID),
  Value = ifelse(ID == 3, 3000, Value)
)

This code replaces the ID where ID is 1, and Value is 10000 and replaces Value where ID is 3.

D. Vectorized Replacement Using dplyr

When dealing with replacements across multiple levels or categories, vectorized solutions are efficient.

library(dplyr)
df <- df %>% mutate(Value = recode(Value, `10000` = 100000, `3000` = 30000))

This example uses the recode function from dplyr to replace 10000 with 100000 and 3000 with 30000 in the Value column of the dataframe df.

E. Handling NA Replacements

In situations where you want to replace NA values, coalesce() from dplyr is handy.

df$Value <- dplyr::coalesce(df$Value, 9999)

This replaces any NA in Value column with 9999.

Conclusion:

Replacing values in R is a fundamental step in data cleaning and preprocessing. Depending on your needs, various functions and packages are available, such as base R functionalities like subsetting and assignment, ifelse(), or more specialized functions from packages like dplyr and plyr.

The choice of method depends on the context, the complexity of the conditions, and personal preference. Understanding the various ways to replace values in R will enable you to manage and manipulate your datasets effectively, preparing them accurately for subsequent analysis.