← SASnRCharacter Functions

Introduction

Character functions cover the everyday chores of clinical string handling: measure a value, count a separator, extract a piece, replace a token, check for missing.
Most SAS character functions have a direct analogue in R, usually in the stringr package plus a few base R helpers.
SAS offers length, countw, scan, countc, find, substr, catx, translate, tranwrd, and missing as the daily toolkit.
Tidyverse mirrors these with str_length, str_count, word, str_locate, str_sub, str_c, chartr, str_replace_all, and is.na.
Base R covers them with nchar, strsplit, gregexpr, regexpr, substr, paste, chartr, gsub, and is.na.
This lesson applies each function to a column of ISO-8601 partial dates so we can compare the outputs side by side across the three languages.

Character Functions

SAS (Base SAS)+

data alldata;
    infile cards dsd dlm='|' truncover;
    input dtc : $20.;
cards;
2025
2025-01
2025-01-10
2025-02-28
2025-03
2025-12-31
2025-12-31T23
2025-12-31T23:59
2025-12-31T23:59:59
.
;
run;

data dates01;
    set alldata;

    length=length(dtc);
    countw=countw(dtc,'T');
    scan_fword=scan(dtc,1,'T');
    countc=countc(dtc,'-');
    findt=find(dtc,'T','i');
    substr_date=substr(dtc,1,10);
    concat=catx(" ", "Event start:",dtc);
    translate=translate(dtc,'-',':');
    tranwrd=tranwrd(dtc,'2025','20xx');
    missing=missing(dtc);

run;

This lesson demonstrates the most common SAS character functions by applying them to a column of ISO-8601 partial dates.
LENGTH(dtc) returns the number of characters in the value — useful for spotting truncated or oversized strings.
COUNTW(dtc, 'T') counts words when T is treated as a delimiter — so "2025-12-31T23:59" has two words.
SCAN(dtc, 1, 'T') grabs the first word with the same delimiter — the date portion before the T.
COUNTC(dtc, '-') counts occurrences of the dash character — handy for distinguishing year-only, year-month and full date formats.
FIND(dtc, 'T', 'i') returns the character position of the T separator, or 0 if absent.
SUBSTR(dtc, 1, 10) extracts the first 10 characters — the full date portion when present.
CATX(" ", "Event start:", dtc) joins values with a space separator, quietly skipping empty pieces.
TRANSLATE(dtc, '-', ':') swaps every colon with a dash, character-by-character.
TRANWRD(dtc, '2025', '20xx') replaces a whole word or substring with another — the go-to for string redaction.
MISSING(dtc) returns 1 if the value is missing and 0 otherwise — safer than testing dtc = "".
After running, we inspect dates01 and confirm each new column reflects the transformation its name suggests.

R (tidyverse)+

alldata <- tribble(
  ~dtc,
  "2025",
  "2025-01",
  "2025-01-10",
  "2025-02-28",
  "2025-03",
  "2025-12-31",
  "2025-12-31T23",
  "2025-12-31T23:59",
  "2025-12-31T23:59:59",
  NA
)


dates01 <- alldata %>%
  mutate(
    length = str_length(dtc),                         # LENGTH(dtc)
    countw = str_count(dtc, "T") + 1,                  # COUNTW(dtc, 'T')
    scan_fword = word(dtc, 1, sep = "T"),             # SCAN(dtc, 1, 'T')
    countc = str_count(dtc, "-"),                     # COUNTC(dtc, '-')
    findt = str_locate(dtc, "T")[, 1],                # FIND(dtc, 'T')
    substr_date = str_sub(dtc, 1, 10),                # SUBSTR(dtc, 1, 10)
    concat = str_c("Event start:", dtc, sep = " "),   # CATX(" ", ...)
    translate = chartr(":","-",  dtc),                # TRANSLATE(dtc, '-', ':')
    tranwrd = str_replace_all(dtc, "2025", "20xx"),   # TRANWRD(dtc, '2025', '20xx')
    missing = is.na(dtc)                              # MISSING(dtc)
  )

The tidyverse equivalents mostly live in stringr, with a couple of base R helpers mixed in.
str_length(dtc) mirrors SAS LENGTH() — character count of a string, NA-safe.
For "count of words" with a separator, we use str_count(dtc, "T") + 1. str_count() counts the separator; adding one gives the word count SAS COUNTW returns.
word(dtc, 1, sep = "T") extracts the first word by a separator — the match for SAS SCAN(dtc, 1, 'T').
str_count(dtc, "-") counts occurrences of a character or pattern — equivalent to SAS COUNTC.
str_locate(dtc, "T")[, 1] returns the start position of the first match — the tidyverse analogue of SAS FIND(). We subscript [, 1] to get the start column only.
str_sub(dtc, 1, 10) mirrors SUBSTR(dtc, 1, 10).
str_c("Event start:", dtc, sep = " ") concatenates with a separator — the equivalent of SAS CATX, though it does not skip NAs automatically (watch for that).
For character-by-character swap, we use base R chartr(":", "-", dtc) — matches SAS TRANSLATE().
str_replace_all(dtc, "2025", "20xx") replaces a substring everywhere it occurs, the same as SAS TRANWRD().
is.na(dtc) is the base R check for missing values — mirrors SAS MISSING() and returns a logical.
After running, we inspect dates01 and confirm each column matches its SAS counterpart row-by-row.

Happy learning!

Search lessons

Introduction

Character Functions