String Manipulation in R

Analogy categoryAnalogies

Working with Strings: Base R vs stringr

Base R provides string functions like paste(), paste0(), nchar(), substr(), toupper(), and tolower(), but their argument order and behavior are inconsistent across functions; the stringr package (part of the tidyverse) wraps most string operations behind a consistent str_ prefix where the string is always the first argument, for example str_length(), str_sub(), str_to_upper(), and str_detect(), making functions easier to chain with the pipe and to remember.

🏏

Cricket analogy: Base R's inconsistent string functions are like a scorecard app where run rate, strike rate, and economy rate each use a different input order, while stringr's str_ prefix is like a redesigned app where every stat function takes the player's name first, consistently.

Pattern Matching with Regular Expressions

Regular expressions describe patterns within text: str_detect(x, '^R') tests whether each string starts with R, str_extract(x, '[0-9]+') pulls out the first run of digits, and str_replace_all(x, '[aeiou]', '*') replaces every vowel; base R's grepl(), grep(), sub(), and gsub() perform the equivalent operations but take the pattern as the first argument instead of the string, matching base R's less consistent convention.

🏏

Cricket analogy: str_detect(x, '^Century') flagging every match report that starts with Century is like a scanner searching newspaper archives for headlines announcing a hundred, pulling out only the relevant articles.

library(stringr)

titles <- c('Intro to R Lv3', 'Advanced R Lv12', 'Data Wrangling Lv7')

str_extract(titles, '\\d+')
str_detect(titles, '^Advanced')
str_replace(titles, 'Lv(\\d+)', 'Level \\1')

str_to_lower(titles) |>
  str_split(' ')

Splitting, Joining, and Case Manipulation

str_split(x, pattern) breaks each string into a list of pieces wherever the pattern matches (e.g. splitting a,b,c on a comma), while str_c() (or base paste0()) joins pieces back together, optionally with a collapse argument to flatten a vector into one string; str_to_upper(), str_to_lower(), and str_to_title() normalize case, which matters because string comparisons and joins in R are case-sensitive by default, so India and india are treated as different values unless case is normalized first.

🏏

Cricket analogy: str_split('India,Australia,England', ',') breaking a comma-separated squad list into individual team names is like a scorer separating a combined fixture list into individual matches to log each one.

Trimming, Padding, and Formatting Strings

str_trim() removes leading and trailing whitespace that often creeps in from manual data entry or copy-pasting (a value like ' R ' becomes R), str_pad() adds characters to reach a fixed width (useful for aligning IDs like zero-padding: str_pad('7', width = 3, pad = '0') gives 007), and glue::glue() offers readable string interpolation with {} placeholders as a more legible alternative to paste0() or sprintf() for building formatted messages from variables.

🏏

Cricket analogy: str_trim() cleaning up a name entered with stray spaces is like a scorer proofreading a scorecard before publishing, while str_pad('7', width = 3, pad = '0') formatting a jersey number as 007 mirrors squad-list conventions.

The glue package's glue('Player {name} scored {runs} runs') interpolates variables directly inside {} braces, avoiding the awkward quote-and-comma juggling of paste0('Player ', name, ' scored ', runs, ' runs').

Regex metacharacters like the dot, parentheses, brackets, plus sign, and dollar sign have special meaning in pattern strings; to match them literally (e.g. a literal period in a filename), escape them with a double backslash in R string literals, such as str_detect(x, '\\.') to detect an actual dot.

Base R string functions (paste0, substr, grepl) work but have inconsistent argument order; stringr's str_ prefix functions are more consistent, always taking the string first.
str_detect(), str_extract(), and str_replace_all() use regular expressions to find, pull out, or replace patterns in text.
str_split() breaks strings into pieces on a pattern; str_c()/paste0() join pieces back together.
String comparisons in R are case-sensitive by default, so normalize case with str_to_upper()/str_to_lower() before comparing or joining.
str_trim() removes stray leading/trailing whitespace; str_pad() pads strings to a fixed width, e.g. for zero-padded IDs.
glue() offers readable {}-based string interpolation as an alternative to chaining paste0() calls.
Regex metacharacters need escaping with a double backslash in R string literals to be matched literally.