Formats a number (or bigint, or string that appears to be a number) as one would see it written in imperial measurements, e.g. "1 1/2" instead of "1.5".

Full documentation

Features:

• To use vulgar fraction characters like "⅞", pass true as the second argument. Other options like Roman numerals are described below.
• String inputs are parsed with numeric-quantity, so mixed numbers ("1 1/2"), vulgar fractions ("½"), bare fractions ("1/3"), and comma- or underscore-separated numbers ("1,000") are all accepted in addition to plain decimal strings.
• The return value will be null if the first argument is not a number, string, or bigint, or is a string with no recognizable numeric portion at the start. Trailing junk is tolerated, so formatQuantity("1.5 cups") returns "1 1/2".
• The return value will be an empty string ("") if the first argument is 0, 0n, or "0", which fits the primary use case of formatting recipe ingredient quantities.
• Values too large or too small for positional notation are returned in JavaScript's exponential form (e.g. formatQuantity(1e21) is "1e+21"), and Infinity/-Infinity are stringified as-is. bigint inputs are never exponential; they stringify in full.

For the inverse operation—converting a string to a number—check out numeric-quantity. It handles mixed numbers, vulgar fractions, comma/underscore separators, and Roman numerals.

If you're interested in parsing recipe ingredient strings, try parse-ingredient.

Usage

Installed

import { formatQuantity } from 'format-quantity';

formatQuantity(1.5); // "1 1/2"
formatQuantity(2.66); // "2 2/3"
formatQuantity(3.875, true); // "3⅞"

CDN

As an ES module:

<script type="module">
  import { formatQuantity } from 'https://cdn.jsdelivr.net/npm/format-quantity/+esm';

  console.log(formatQuantity(10.5)); // "10 1/2"
</script>

As UMD (all exports are properties of the global object FormatQuantity):

<script src="https://unpkg.com/format-quantity"></script>
<script>
  console.log(FormatQuantity.formatQuantity(10.5)); // "10 1/2"
</script>

Options

The second parameter to formatQuantity can be a boolean value or an options object.

vulgarFractions

| Type | Default | | --------- | ------- | | boolean | false |

Returns vulgar fractions when appropriate. This option has the same effect as passing a plain boolean value as the second parameter.

formatQuantity(3.875, { vulgarFractions: true }); // "3⅞"
// is the same as
formatQuantity(3.875, true); // "3⅞"

Note: formatQuantity supports sixteenths, but no vulgar fraction characters exist for that denomination. Therefore the vulgarFractions option has no effect if the fraction portion of the final string is an odd numerator over a denominator of 16.

fractionSlash

| Type | Default | | --------- | ------- | | boolean | false |

Uses the fraction slash character ("\u2044") to separate the numerator and denominator instead of the regular "solidus" slash ("\u002f"), with Unicode superscript numerator and subscript denominator digits. This option is ignored if the vulgarFractions option is also true.

formatQuantity(3.875, { fractionSlash: true }); // "3 ⁷⁄₈"
formatQuantity(3.875, { fractionSlash: true, vulgarFractions: true }); // "3⅞"

separator

| Type | Default | | -------- | ------- | | string | N/A |

Overrides the string placed between the whole number and the fraction. When not specified, the default is " " (a space) for ASCII and fraction-slash fractions, and "" (no space) for vulgar fractions. Common alternatives include a hyphen ("-") and a no-break space ("\u00a0").

formatQuantity(1.5, { separator: '-' }); // "1-1/2"
formatQuantity(1.5, { separator: ' ', vulgarFractions: true }); // "1 ½"
formatQuantity(1.5, { separator: '\u00a0' }); // "1\u00a01/2" (no-break space)

tolerance

| Type | Default | | ----------------- | -------: | | number \| false | 0.0075 |

This option determines how close the decimal portion of a number has to be to the actual quotient of a fraction to be considered a match. For example, consider the fraction 1⁄3: $1 \div 3 = 0.\overline{333}$, repeating forever. The number 0.333 (exactly 333 thousandths) is not equivalent to 1⁄3, but it's very close. So even though $0.333 \neq 1 \div 3$, both formatQuantity(0.333) and formatQuantity(1/3) will return "1/3".

The window is centered on the exact quotient and is checked against every candidate fraction; when more than one is within the window, the closest one wins regardless of evaluation order.

A lower tolerance increases the likelihood that formatQuantity will return a decimal representation instead of a fraction or mixed number since the matching algorithm will be stricter. A higher tolerance increases the likelihood that formatQuantity will return a fraction or mixed number, but at the risk of matching a fraction that is only loosely related to the input.

// Low tolerance - returns a decimal since 0.333 is not close enough to 1/3
formatQuantity(0.333, { tolerance: 0.00001 }); // "0.333"
// High tolerance - 0.3 is within 0.1 of both 5/16 and 1/3, and 5/16 is closer
formatQuantity(0.3, { tolerance: 0.1 }); // "5/16"

Two values are special:

• 0 means only exact quotients match. Anything else falls through to its decimal representation.
• false disables fraction matching entirely, so every non-integer is returned as a decimal.

formatQuantity(1.5, { tolerance: 0 }); // "1 1/2" (0.5 is exactly 1 ÷ 2)
formatQuantity(1.51, { tolerance: 0 }); // "1.51"
formatQuantity(1.5, { tolerance: false }); // "1.5"

Any other value—a negative number, NaN, a numeric string, null, undefined—is ignored and the default is used instead.

zeroFormat

| Type | Default | | -------- | ------- | | string | "" |

Specify the string to return when the input numerically evaluates to zero (0).

formatQuantity(0, { zeroFormat: '' }); // "" (default)
formatQuantity(0, { zeroFormat: '0' }); // "0"
formatQuantity(0, { zeroFormat: 'N/A' }); // "N/A"

allowTrailingInvalid

| Type | Default | | --------- | ------- | | boolean | true |

If input is a string, ignore trailing non-numeric input (à la parseFloat). Set to false for strict parsing — any invalid characters will result in null.

formatQuantity('123abc', { allowTrailingInvalid: true }); // "123"
formatQuantity('123abc', { allowTrailingInvalid: false }); // null

romanNumerals

| Type | Default | | --------- | ------: | | boolean | false |

Coerces the number into an integer using Math.floor, then formats the value as Roman numerals. The algorithm uses strict, modern rules, so the number must be between 1 and 3999 inclusive (or between 1n and 3999n — bigint is also allowed). Values outside that range return null.

When this option is true, all other options are ignored.

formatQuantity(1214, { romanNumerals: true }); // "MCCXIV"
formatQuantity(12.14, { romanNumerals: true, vulgarFractions: true }); // "XII"
formatQuantity(4000, { romanNumerals: true }); // null
formatQuantity(-1, { romanNumerals: true }); // null

formatQuantity(0, …) returns "" (or the configured zeroFormat) regardless of this option, since the zero rule is applied before options are processed.