← SASnRApply Basic Numeric Functions

Introduction

Basic numeric functions — round, ceil, floor, truncate, absolute value, modulo — show up everywhere in data manipulation, from banding continuous scores to building derived flags.
Most SAS numeric functions have direct analogues in R, with one notable exception: the sign convention for the modulo operator differs between SAS and R.
SAS offers ROUND, CEIL, FLOOR, INT, ABS, and MOD.
R offers round, ceiling, floor, trunc, abs, and %% (with a small helper needed to match SAS MOD).
This lesson applies each function to a vector that includes negative values, a zero, and positive fractions, so edge cases are easy to see.
Pay particular attention to the two modulo columns — they only diverge on negative numerators, which is exactly where the two conventions disagree.

Apply Basic Numeric Functions

SAS (Base SAS)+

data numbers;
input num;
round_val = round(num);
ceil_val = ceil(num);
floor_val = floor(num);
int_val = int(num);
abs_val = abs(num);
mod_val = mod(num, 3);
datalines;
-2.51
-1.2
0
1.3
2.7
;
run;

This lesson walks through the most common SAS numeric functions applied to a small vector of floats.
ROUND(num) rounds to the nearest whole number using half-up banker rounding.
CEIL(num) rounds up towards positive infinity — always gives the next integer at or above the value.
FLOOR(num) rounds down towards negative infinity — always gives the next integer at or below the value.
INT(num) truncates towards zero — drops the fractional part without rounding.
ABS(num) returns the magnitude, always non-negative.
MOD(num, 3) returns the remainder after integer division by 3, with the same sign as the numerator.
After running, we inspect the numbers dataset and confirm each derived column reflects the expected rounding or remainder rule.
Useful for building derived flags (for example, rounding a continuous age to a whole number) or for arithmetic grouping.

R (tidyverse)+

mod_sas_style <- function(x, y) {
  x - y * trunc(x / y)
}

df <- tibble(
  num = c(-2.51, -1.2, 0, 1.3, 2.7)
) %>%
  mutate(
    round_val = round(num),
    ceil_val  = ceiling(num),
    floor_val = floor(num),
    int_val   = trunc(num),
    abs_val   = abs(num),

mod_val1   = num%%3, 
    mod_val2   = mod_sas_style(num, 3)  
  )

Most of the SAS numeric functions have direct base R equivalents; tidyverse pipelines them inside mutate().
round(num) matches SAS ROUND() — although R uses banker's rounding (round-half-to-even), which differs subtly from SAS for .5 values.
ceiling(num) is the R equivalent of SAS CEIL().
floor(num) matches SAS FLOOR() exactly.
trunc(num) truncates towards zero, matching SAS INT().
abs(num) is identical to SAS ABS().
For modulo, R has two options. num %% 3 is R's built-in modulo, but it follows the mathematical convention (result has the sign of the divisor). SAS MOD() uses the truncation convention (sign of the numerator).
We define a small helper mod_sas_style(x, y) = x - y * trunc(x / y) to reproduce the SAS behaviour exactly, and include both for comparison.
After running, we confirm mod_val1 (R convention) and mod_val2 (SAS convention) only disagree on negative inputs — exactly where the two conventions diverge.

Happy learning!

Search lessons

Introduction

Apply Basic Numeric Functions