Capitalize with dplyr

Question

I am doing data cleaning with dplyr. One of the things I want to do is to capitalize values in certain columns.

    data$surname
    john
    Mary
    John
    mary
    ...

I suppose I have to use the mutate function of dplyr with something like this

    titleCase <- function(x) {
    + s <- strsplit(as.character(x), " ")[[1]]
    + paste(toupper(substring(s, 1, 1)), substring(s, 2),
    + sep = "", collapse = " ")
    + }

But how to combine both? I get all kinds of errors or truncated data frames

Thanks

akrun · Answer 1 · 2016-05-11T11:19:37.217

9

We can use sub

sub("(.)", "\\U\\1", data$surname, perl=TRUE)
#[1] "John" "Mary" "John" "Mary"

Implementing in the dplyr workflow

library(dplyr)
data %>%
     mutate(surname = sub("(.)", "\\U\\1", surname, perl=TRUE))

If we need to do this on multiple columns

data %>%
     mutate_each(funs(sub("(.)", "\\U\\1", ., perl=TRUE)))

Just to check

res <- data1 %>%  
          mutate(surname = sub("(.)", "\\U\\1", surname, perl=TRUE))
sum(grepl("[A-Z]", substr(res$surname, 1,1)))
#[1] 500000

data

data <- data.frame(surname=c("john", "Mary", "John", "mary"), 
firstname = c("abe", "Jacob", "george", "jen"), stringsAsFactors=FALSE)

data1 <-  data.frame(surname = sample(c("john", "Mary", "John", "mary"), 
    500000, replace=TRUE), stringsAsFactors=FALSE)

edited May 11 '16 at 11:19

answered May 11 '16 at 09:43

akrun

874,273
37
540
662

Thanks Akrun, It works on a reasonably sized data frame, I've checked. But no effect, and no error displayed either, on my 500000 rows data frame. – Philippe Ramirez May 11 '16 at 10:27
@PhilippeRamirez That's strange. How many columns you were changing? – akrun May 11 '16 at 10:28
One at a time, with your first dplyr function – Philippe Ramirez May 11 '16 at 10:33
@PhilippeRamirez I couldn't reproduce the problem. Please check the update – akrun May 11 '16 at 11:19
Ok, the pb would be with the visualization. Applying the function on a small table will show me the full results (on RStudio) with correct capitalization. But a head(data) will return the original spellings. And similarly with a write.csv. What should I do to capture the results nicely? – Philippe Ramirez May 11 '16 at 12:29
@PhilippeRamirez I don't use `RStudio`. So I can't comment on it. – akrun May 11 '16 at 12:47
I found a way: assign the results to a new table. Thanks – Philippe Ramirez May 11 '16 at 12:57
@PhilippeRamirez That is what I showed in `res1 <- ...` If you need it to be in place, try `library(magrittr);data %<>% mutate(surname = sub("(.)", "\\U\\1", surname, perl=TRUE))` It will replace the values in the original 'data'. – akrun May 11 '16 at 12:58

score 8 · Answer 2 · answered Feb 16 '20 at 02:37

A little late to the party but you can use stringr package

library(stringr)
library(dplyr)

example1 <- tibble(names = c("john" ,"Mary", "John", "mary"))

example1 %>%
mutate(names = str_to_title(names))

##  names
##  <chr>
## 1 John 
## 2 Mary 
## 3 John 
## 4 Mary

This will still work if you want all terms capitalized

example2 <- tibble(names = c("john james" ,"Mary carey", "John Jack", "mary Harry"))

example2 %>%
mutate(names = str_to_title(names))

##  names
##  <chr>
## 1 John James 
## 2 Mary Carey 
## 3 John Jack
## 4 Mary Harry

If you only want the first term capitalized, str_to_sentence() will work

example2 %>%
mutate(names = str_to_sentence(names))

##  names
##  <chr>
## 1 John james 
## 2 Mary carey 
## 3 John jack
## 4 Mary harry

RHertel · Answer 3 · 2016-05-11T10:19:18.153

6

There is a dedicated function for this that you can try:

R.utils::capitalize(data$surname)

If this needs to be implemented into a dplyr procedure, one could try the following:

library(dplyr)
library(R.utils)
data %>% mutate(surname = capitalize(surname))

edited May 11 '16 at 10:19

answered May 11 '16 at 09:45

RHertel

23,412
5
38
64

Thanks RHertel. Your solution seems to work also. My pb is that I cannot get the final product. As I reponded to Akrun above, applying the function on a small table will show me the full results (on RStudio) with correct capitalization. But a head(data) will return the original spellings. And similarly with a write.csv. What should I do to capture the results nicely? – Philippe Ramirez May 11 '16 at 12:31
I found a way: assign the results to a new table. Thanks – Philippe Ramirez May 11 '16 at 12:57

Capitalize with dplyr

3 Answers3

data