joeroe / numerals

R package for numeric vectors with Unicode numerals

numerals

numerals is an R package that extends the base numeric types with methods for printing using Unicode digits in other numeral systems. It currently supports Eastern Arabic ("ar"), Bengali ("bn"), Persian ("fa"), and Burmese ("my") numerals.

Installation

You can install the development version of numerals from GitHub using devtools:

# install.package("devtools")
devtools::install_github("joeroe/numerals")

numerals has not yet been released on CRAN.

Usage

numeral() creates a numeric vector which is printed in another numeral system:

library("numerals")
# Eastern Arabic numerals
numeral(1:10, "ar")
#> <numeral[10]>
#>  [1]  ١  ٢  ٣  ٤  ٥  ٦  ٧  ٨  ٩ ١٠

# Persian numerals
numeral(1:10, "fa")
#> <numeral[10]>
#>  [1]  ۱  ۲  ۳  ۴  ۵  ۶  ۷  ۸  ۹ ۱۰

Numerals are compatible with both base and tidyverse packages. For example, they are printed in data frames and tibbles:

library("tibble")
x <- data.frame(en = 1:10, ar = numeral(1:10, "ar"), fa = numeral(1:10, "fa"))

x
#>    en ar fa
#> 1   1  ١  ۱
#> 2   2  ٢  ۲
#> 3   3  ٣  ۳
#> 4   4  ٤  ۴
#> 5   5  ٥  ۵
#> 6   6  ٦  ۶
#> 7   7  ٧  ۷
#> 8   8  ٨  ۸
#> 9   9  ٩  ۹
#> 10 10 ١٠ ۱۰
as_tibble(x)
#> # A tibble: 10 × 3
#>       en     ar     fa
#>    <int> <numr> <numr>
#>  1     1      ١      ۱
#>  2     2      ٢      ۲
#>  3     3      ٣      ۳
#>  4     4      ٤      ۴
#>  5     5      ٥      ۵
#>  6     6      ٦      ۶
#>  7     7      ٧      ۷
#>  8     8      ٨      ۸
#>  9     9      ٩      ۹
#> 10    10     ١٠     ۱۰

Otherwise, numerals are freely coercible to base numerics and so can be used in calculations:

numeral(2, "ar") * 10
#> <numeral[1]>
#> [1] ٢٠

numerals's Issues

Compatibility with base plots

Right-to-left format method

R seems to have some support for right-to-left scripts, presumably picking up on the unicode bidi class, e.g.:

library("numerals")
numeral(1:9, "ar")
#> <numeral[9]>
#> [1] ١ ٢ ٣ ٤ ٥ ٦ ٧ ٨ ٩

But it doesn't always happen when expected:

numeral(1:9, "fa")
#> <numeral[9]>
#> [1] ۱ ۲ ۳ ۴ ۵ ۶ ۷ ۸ ۹

And it would also be nice if the indices and justification followed, something like this:

library("numerals")
numeral(1:9, "ar")
#> <numeral[9]>
#> ١] ١ ٢ ٣ ٤ ٥ ٦ ٧ ٨ ٩]

Write vignette

Including:

• numeral class
• Plotting with numerals
• Reading/writing numerals
  • base write.* and write_* seem to work fine out of the box

numr_replace() or numr_substitute()?

Which to use as the default for formatting numerals? numr_replace() uses string replacement; numr_substitute() decomposes a numeric into digits and directly subsets the numerals matrix.

I was hoping numr_substitute() would be faster, but it isn't (yet):

library("numerals")

x <- floor(runif(1000) * 1e6)

bench::mark(
  numr_substitute(x, "fa")
)
#> Warning: Some expressions had a GC in every iteration; so filtering is disabled.
#> # A tibble: 1 x 6
#>   expression                    min   median `itr/sec` mem_alloc `gc/sec`
#>   <bch:expr>               <bch:tm> <bch:tm>     <dbl> <bch:byt>    <dbl>
#> 1 numr_substitute(x, "fa")   18.9ms     21ms      45.2        NA     68.7

bench::mark(
  numr_replace(x, "fa")
)
#> # A tibble: 1 x 6
#>   expression                 min   median `itr/sec` mem_alloc `gc/sec`
#>   <bch:expr>            <bch:tm> <bch:tm>     <dbl> <bch:byt>    <dbl>
#> 1 numr_replace(x, "fa")   8.51ms   8.54ms      112.        NA        0

But I do suspect numr_substitute() is more robust to e.g. character encoding and locale issues.

Compatability with ggplot2

Add functions for reading data in different numeral systems

Infer default numeral system from locale

This would be convenient for interactive use, though for portability we should recommend explicitly specifying the system.

Decimal separators

Add numeral systems

From https://en.wikipedia.org/wiki/List_of_numeral_systems:

• Proto-cuneiform numerals
• Proto-Elamite numerals
• Sumerican numerals
• Egyptian numerals
• Elamite numerals
• Indus numerals
• Babylonian numerals
• Chinese numerals
• Japanese numerals
• Korean numerals (Sino-Korean)
• Vietnamese numerals (Sino-Vietnamese)
• Aegean numerals
• Bengali numerals
• Roman numerals
• Hebrew numerals
• Indian numerals
  • Tamil numerals
  • Devanagari numerals
  • Tibetan numerals
• Greek numerals
• Phoenician numerals
• Chinese rod numerals
• Ge'ez numerals
• Armenian numerals
• Khmer numerals
• Thai numerals
• Abjad numerals
• Eastern Arabic numerals
• Persian numerals
• Vietnamese numerals (Chữ Nôm)
• Western Arabic numerals
• Glagolitic numerals
• Cyrillic numerals
• Burmese numerals
• Tangut numerals
• Cistercian numerals
• Maya numerals
• Muisca numerals
• Korean numerals (Hangul)
• Aztec numerals
• Sinhala numerals
• Pentimal runes
• Kaktovik Inupiaq numerals

Consider replacing ISO language codes with unicode script codes

In numeral(system = "<iso code>"). https://www.unicode.org/iso15924/iso15924-codes.html