How to replace an exact string using stringr functions?

Question

I'm trying to replace exact strings in a column using stringr functions.

The dataset I try it on is this:

data <- data.frame(
  column = c("Value", "Values", "Value", "Values")
)

data

  column
1 Value
2 Values
3 Value
4 Values

I want to replace "Value" with "Values". I tried str_replace(data$column, "Value", "Values"), but this creates the following unwanted replacements:

[1] "Values"  "Valuess" "Values"  "Valuess"

I'd like the output to be:

[1] "Values"  "Values" "Values"  "Values"

try `str_replace(data$column, "Value\\b", "Values")` and have a look [here](https://stackoverflow.com/questions/7227976/using-grep-in-r-to-find-strings-as-whole-words-but-not-strings-as-part-of-words). — Roman, Aug 02 '18 at 07:33
as suggested by @Ravinder. Please use `sub()` for this task. Stay `base::` where you can. — Andre Elrico, Aug 02 '18 at 08:17
Can you explain a little bit why one should choose base where one can? — MetaPhilosopher, Aug 02 '18 at 13:02

score 5 · Accepted Answer · answered Aug 02 '18 at 07:44

5

Here are a few possibilities using regular expressions:

x <- c("value", "values")
str_replace(x, "value$", "values") #method 1
str_replace(x, "value\\b", "values") #method 2
str_replace(x, "value(?!s)", "values") #method 3

all of the above return the same

[1] "values" "values"

A short explanation: the first method looks for 'value' at the end of a string. The symbol $ matches the end of the string.

The second method looks for 'value' followed by a word boundary.

The third method looks for 'value' followed by anything but the symbol 's'.

You can find a helpful cheat sheet about stringr and regular expressions here. Hope this helps.

answered Aug 02 '18 at 07:44

Cettt

11,460
7
35
58

Thank you! Do you have any suggested comprehensive resources to learn more about regular expressions in stringr context? – MetaPhilosopher Aug 02 '18 at 07:55
2

I personally struggled a lot to understand regular expressions. What helped me in the end was the second page of the cheat sheet I posted in the answer plus a couple of hours experimenting with the examples they provide there. – Cettt Aug 02 '18 at 07:57
1

@MetaPhilosopher regex101.com – Andre Elrico Aug 02 '18 at 08:36
Use `^value$` for start and end of string. – qwr Apr 25 '23 at 01:12

score 2 · Answer 2 · answered Aug 02 '18 at 07:33

2

Just a simple string comparison should do the trick.

data[data$col == "Value","col"] = "Values"

answered Aug 02 '18 at 07:33

MSW Data

441
3
8

RavinderSingh13 · Answer 3 · 2018-08-02T08:00:36.877

Could you please try following:

sub("Value[a-z]+","Values",data$column)

Output will be as follows.

sub("Value[a-z]+","Values",data$column,perl = TRUE)
[1] "Values" "Values" "Values" "Values"

Explanation: Following is only for explanation purposes.

sub(             ##using sub function of R whose method is: sub(regex_to_match_in_current_value,new_value_which_should_be_there_after_match,variable)
"Value[a-z]+",   ##mentioning Value string with [a-z]+ alphabets till their regular sequences.
"Value",         ##Substitute above match of strings with only string Value here.
data$column)     ##Mentioning data frame data with its column.

Where sample data is from:

data <- data.frame(
  column = c("Value", "Values", "Value", "Values")
)

score 0 · Answer 4 · answered Dec 05 '20 at 19:10

0

data$column <- ifelse(data$column=='Value','Values','Values')

answered Dec 05 '20 at 19:10

Mumtaj Ali

421
4
7

How to replace an exact string using stringr functions?

4 Answers4

Linked