paste() and paste0() in R: Concatenate Strings

The paste() and paste0() functions in base R combine strings or vectors into a single string or character vector. paste0 is a shortcut for paste(..., sep = "").

paste("Hello", "world")                       # "Hello world" (default sep = " ")
paste0("Hello", "world")                      # "Helloworld" (no separator)
paste("a", "b", sep = "-")                    # "a-b"
paste(c("a","b","c"), collapse = ",")         # "a,b,c" (one string)
paste(c("a","b"), c("1","2"), sep = "_")      # vectorized: "a_1" "b_2"
paste(c("a","b"), c("1","2"), collapse = "-") # "a 1-b 2"
sprintf("%s is %d", "Alice", 30)              # alternative for templating

What paste() does in one sentence

paste(...) takes any number of strings or vectors and combines them, controlled by sep (between inputs) and collapse (between elements of the result vector). Without collapse, the output is a vector; with collapse, the result is a single string.

This is the most-used string concatenation function in base R. Every R programmer learns it early. paste0() is the no-separator shortcut.

Syntax

paste(..., sep = " ", collapse = NULL). Variadic arguments are recycled to the longest length.

Five common patterns

1. Two strings into one

2. Custom separator

3. Vectorized concatenation

Pairs of elements are joined. Recycling rules apply if the lengths differ.

4. Collapse vector to single string

collapse = ", " joins all elements of the result with the specified separator into ONE string.

5. Combine sep and collapse

sep joins INPUT vectors element-wise. collapse joins the resulting vector into one string.

paste() vs sprintf() vs glue::glue() vs str_c()

Four ways to build strings in R, each with different strengths.

When to use which:

• Use paste0 for simple "join these strings".
• Use sprintf for fixed-width / formatted output.
• Use glue when you embed variables in long template strings.
• Use str_c if you need predictable NA behavior in tidyverse pipelines.

Common pitfalls

Pitfall 1: paste with NA propagates. paste("Alice", NA) returns "Alice NA" (the NA is converted to "NA"). Use str_c() or filter NAs first if this is wrong.

Pitfall 2: forgetting collapse. paste(c("a","b")) returns the SAME vector unchanged (length 2). To get a single string, add collapse = "".

Try it yourself

Related string functions

After mastering paste, look at:

• sprintf(): format with %s, %d, %f, etc.
• glue::glue(): template-based interpolation ({var} syntax)
• stringr::str_c(): stringr's vectorized concatenation
• format(): format numbers with specific width and digits
• file.path(): build file paths with the right OS separator
• tidyr::unite(): data frame column-wise paste

For embedding variables in long template strings, glue("Hello {name}, you are {age}") is much cleaner than paste0.

FAQ

What is the difference between paste and paste0 in R?

paste0(...) is paste(..., sep = ""): it concatenates with NO separator. paste(...) defaults to sep = " ". Use paste0 when you do not want a space (or any separator); paste with sep otherwise.

How do I concatenate strings in R?

paste0("Hello", "world") for no separator. paste("Hello", "world", sep = "-") for a custom separator. For collapsing a vector to one string: paste(c("a","b"), collapse = "-").

What is the difference between sep and collapse in paste?

sep joins MULTIPLE INPUTS element-wise (e.g., paste(c("a","b"), c("1","2"), sep = "_") returns c("a_1","b_2")). collapse joins the RESULT VECTOR into a single string (e.g., paste(..., collapse = ",") returns "a,b").

How do I paste with formatting like sprintf?

Use sprintf("%s is %d years old", name, age) for formatted output. paste does not interpolate variables into a template the way sprintf does.

Why does paste with NA produce "NA"?

paste converts every input to character; as.character(NA) is "NA" (the string). To skip NAs, use stringr::str_c() (returns NA when any input is NA) or filter NAs before pasting.