[[numeral]]
== Numeral Formatting

Numeral formatting in {kib} is done through a pattern-based syntax.
These patterns express common number formats in a concise way, similar
to date formatting. While these patterns are originally based on Numeral.js,
they are now maintained by {kib}.

Numeral formatting patterns are used in multiple places in {kib}, including:

* <<advanced-options, Advanced settings>>
* <<field-formatters-numeric, Index pattern formatters>>
* <<tsvb, *TSVB*>>
* <<canvas, *Canvas*>>

The simplest pattern format is `0`, and the default {kib} pattern is `0,0.[000]`.
The numeral pattern syntax expresses:

Number of decimal places:: The `.` character turns on the option to show decimal
places using a locale-specific decimal separator, most often `.` or `,`.
To add trailing zeroes such as `5.00`, use a pattern like `0.00`.
To have optional zeroes, use the `[]` characters.
Thousands separator:: The thousands separator `,` turns on the option to group
thousands using a locale-specific separator. The separator is most often `,` or `.`,
and sometimes ` `.
Accounting notation:: Putting parentheses around your format like `(0.00)` will use accounting notation to show negative numbers.

The display of these patterns is affected by the <<kibana-general-settings, advanced setting>> `format:number:defaultLocale`.
The default locale is `en`, but some examples will specify that they are using an alternate locale.

Most basic examples:

|===
| **Input** | **Pattern** | **Locale** | **Output**
| 10000.23 | 0,0 | en (English) | 10,000
| 10000.23 | 0.0 | en (English) | 10000.2
| 10000.23 | 0,0.0 | fr (French) | 10 000,2
| 10000.23 | 0,0.000 | fr (French) | 10 000,230
| 10000.23 | 0,0[.]0 | en (English) | 10,000.2
| 10000.23 | 0.00[0] | en (English) | 10,000.23
| -10000.23 | (0) | en (English) | (10000)
|===

[float]
=== Percentages

By adding the `%` symbol to any of the previous patterns, the value
is multiplied by 100 and the `%` symbol is added in the place indicated.

The default percentage formatter in {kib} is `0,0.[000]%`, which shows
up to three decimal places.

|===
| **Input** | **Pattern** | **Locale** | **Output**
| 0.43 | 0,0.[000]% | en (English) | 43.00%
| 0.43 | 0,0.[000]% | fr (French) | 43,00%
| 1 | 0% | en (English) | 100%
| -0.43 | 0 % | en (English) | -43 %
|===

[float]
=== Bytes and bits

The bytes and bits formatters will shorten the input by adding a suffix like `GB` or `TB`. Bytes and bits formatters include the following suffixes:

`b`:: Bytes with binary values and suffixes. 1024 = `1KB`
`bb`:: Bytes with binary values and binary suffixes. 1024 = `1KiB`
`bd`:: Bytes with decimal values and suffixes. 1000 = `1kB`
`bitb`:: Bits with binary values and suffixes. 1024 = `1Kibit`
`bitd`:: Bits with decimal values and suffixes. 1000 = `1kbit`

Suffixes are not localized with this formatter.

|===
| **Input** | **Pattern** | **Locale** | **Output**
| 2000 | 0.00b | en (English) | 1.95KB
| 2000 | 0.00bb | en (English) | 1.95KiB
| 2000 | 0.00bd | en (English) | 2.00kB
| 3153654400000 | 0.00bd | en (English) | 3.15GB
| 2000 | 0.00bitb | en (English) | 1.95Kibit
| 2000 | 0.00bitd | en (English) | 2.00kbit
|===

[float]
=== Currency

Currency formatting is limited in {kib} due to the limitations of the pattern
syntax. To enable currency formatting, use the symbol `$` in the pattern syntax.
The number formatting locale will affect the result.

|===
| **Input** | **Pattern** | **Locale** | **Output**
| 1000.234 | $0,0.00 | en (English) | $1,000.23
| 1000.234 | $0,0.00 | fr (French) | €1 000,23
| 1000.234 | $0,0.00 | chs (Simplified Chinese) | ¥1,000.23
|===

[float]
=== Duration formatting

Converts a value in seconds to display hours, minutes, and seconds.

|===
| **Input** | **Pattern** | **Output**
| 25 | 00:00:00 | 0:00:25
| 25 | 00:00 | 0:00:25
| 238 | 00:00:00 | 0:03:58
| 63846 | 00:00:00 | 17:44:06
| -1 | 00:00:00 | -0:00:01
|===

[float]
=== Displaying abbreviated numbers

The `a` pattern will look for the shortest abbreviation for your
number, and use a locale-specific display for it. The abbreviations
`aK`, `aM`, `aB`, and `aT` can indicate that the number should be
abbreviated to a specific order of magnitude.

|===
| **Input** | **Pattern** | **Locale** | **Output**
| 2000000000 | 0.00a | en (English) | 2.00b
| 2000000000 | 0.00a | ja (Japanese) | 2.00十億
| -5444333222111 | 0,0 aK | en (English) | -5,444,333,222 k
| -5444333222111 | 0,0 aM | en (English) | -5,444,333 m
| -5444333222111 | 0,0 aB | en (English) | -5,444 b
| -5444333222111 | 0,0 aT | en (English) | -5 t
|===

[float]
=== Ordinal numbers

The `o` pattern will display a locale-specific positional value like `1st` or `2nd`.
This pattern has limited support for localization, especially in languages
with multiple forms, such as German.

|===
| **Input** | **Pattern** | **Locale** | **Output**
| 3 | 0o | en (English) | 3rd
| 34 | 0o | en (English) | 34th
| 3 | 0o | es (Spanish) | 2er
| 3 | 0o | ru (Russian) | 3.
|===

[float]
=== Complete number pattern reference

These number formats, combined with the previously described patterns,
produce the complete set of options for numeral formatting.
The output here is all for the `en` locale.

|===
| **Input** | **Pattern** | **Output**
| 10000 | 0,0.0000 | 10,000.0000
| 10000.23 | 0,0 | 10,000
| -10000 | 0,0.0 | -10,000.0
| 10000.1234 | 0.000 | 10000.123
| 10000 | 0[.]00 | 10000
| 10000.1 | 0[.]00 | 10000.10
| 10000.123 | 0[.]00 | 10000.12
| 10000.456 | 0[.]00 | 10000.46
| 10000.001 | 0[.]00 | 10000
| 10000.45 | 0[.]00[0] | 10000.45
| 10000.456 | 0[.]00[0] | 10000.456
| -10000 | (0,0.0000) | (10,000.0000)
| -12300 | +0,0.0000 | -12,300.0000
| 1230 | +0,0 | +1,230
| 100.78 | 0 | 101
| 100.28 | 0 | 100
| 1.932 | 0.0 | 1.9
| 1.9687 | 0 | 2
| 1.9687 | 0.0 | 2.0
| -0.23 | .00 | -.23
| -0.23 | (.00) | (.23)
| 0.23 | 0.00000 | 0.23000
| 0.67 | 0.0[0000] | 0.67
| 1.005 | 0.00 | 1.01
| 1e35 | 000 | 1e+35
| -1e35 | 000 | -1e+35
| 1e-27 | 000 | 1e-27
| -1e-27 | 000 | -1e-27
|===