Skip to content
Tutorials Center
My learning

Topics

PHP Reference ยท Math Functions

PHP BCMath functions

Reference for PHP BCMath: bcadd, bcsub, bcmul, bcdiv, bcmod, bcpow, bcsqrt, bccomp, scale handling, rounding in PHP 8.4 and the BcMath\Number class.

BCMath is an extension for arbitrary-precision decimal arithmetic. Numbers are passed and returned as strings, so they can have as many digits as memory allows and are free of binary floating-point error. This is the standard way to handle money, tax rates and other values that must add up exactly.

The key concept is scale: the number of digits kept after the decimal point. Every function accepts an optional $scale; if you leave it out, the default scale (0 unless set by bcscale() or the bcmath.scale ini setting) is used. Results are truncated, not rounded, to that scale. The extension must be enabled (extension_loaded("bcmath")).

PHP
<?php
echo 0.1 + 0.2;                 // 0.3 (looks fine, but is 0.30000000000000004)
echo bcadd("0.1", "0.2", 1);    // 0.3, exactly

Try it

PHP Exact arithmetic with a scale

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with fgets(STDIN)), it gets the next line from this box.

PHP Money totals, rounding and errors

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with fgets(STDIN)), it gets the next line from this box.

Syntax

PHP
bcadd(string $num1, string $num2, ?int $scale = null): string
bcsub(string $num1, string $num2, ?int $scale = null): string
bcmul(string $num1, string $num2, ?int $scale = null): string
bcdiv(string $num1, string $num2, ?int $scale = null): string
bcmod(string $num1, string $num2, ?int $scale = null): string
bcpow(string $num, string $exponent, ?int $scale = null): string
bcsqrt(string $num, ?int $scale = null): string
bccomp(string $num1, string $num2, ?int $scale = null): int
bcscale(?int $scale = null): int
bcpowmod(string $num, string $exponent, string $modulus, ?int $scale = null): string
bcfloor(string $num): string      // PHP 8.4
bcceil(string $num): string       // PHP 8.4
bcround(string $num, int $precision = 0, RoundingMode $mode = RoundingMode::HalfAwayFromZero): string  // PHP 8.4
Name Description
$num1, $num2, $num Numbers as strings: optional sign, digits, optional decimal point. No exponent notation, no spaces.
$scale Digits kept after the decimal point. null uses the default scale.
$exponent For bcpow(), an integer-valued string (may be negative).
$modulus For bcpowmod(), the divisor.

Return value: a numeric string with exactly $scale decimals (zero decimals means no decimal point). bccomp() returns -1, 0 or 1, comparing only up to $scale digits. bcscale() returns the previous scale.

Errors: ValueError for a malformed number string or a negative scale; DivisionByZeroError for division or modulus by zero and for bcpowmod() with a zero modulus.

Usage notes

  • Pass strings. Passing floats works, but the float's imprecise value is converted first, defeating the purpose. Keep money as strings from the database onward, or as integer cents.
  • Truncation, not rounding. bcdiv("2", "3", 2) is 0.66. Compute with extra digits and round at the end (bcround() in PHP 8.4, or add half a unit as in the example).
  • Scale 0 by default makes bcadd("1.9", "1.9") return 3. Set bcscale() early or pass the scale each time.
  • Compare with bccomp(), never with == on the strings ("1.0" and "1.00" differ as strings) and not with < on floats.
  • Speed: BCMath is orders of magnitude slower than native ints and floats. Use it for money and exactness, not numerical computing.
  • PHP 8.4 additions: bcfloor(), bcceil(), bcround(), bcdivmod() and the immutable BcMath\Number class, which supports the ordinary operators (+, -, *, /, **, comparison) on exact numbers.
  • Negative zero: since PHP 7.2 results never show -0.
  • Not installed? On some systems it needs enabling; use extension_loaded("bcmath"). For huge integers only, GMP is faster.