Add unit "angstrom"

[singularitti]

I was going to contribute to UnitfulAtomic.jl, but the author asks me to try here first.

I did not find any abbreviation of "angstrom" as "mi" does. But requiring input Å as a unit is harder than just typing angstrom. If you have better ideas, I would like to adopt.

[singularitti]

Wiki for angstrom: https://en.wikipedia.org/wiki/Angstrom

[ajkeller34]

I would suggest adding a line const Å = angstrom so that people who want to type it are able to do so but otherwise I agree this is a well-known and common unit that should probably go here.

Was it necessary to write Unitful.nm in the macro?

[giordano]

Note that Angstrom is already in UnitfulAstro.jl: https://github.com/JuliaAstro/UnitfulAstro.jl/blob/5598f8357944eeb6e5c1885adba444f3cc1f86a9/src/UnitfulAstro.jl#L55

[singularitti]

Note that Angstrom is already in UnitfulAstro.jl: https://github.com/JuliaAstro/UnitfulAstro.jl/blob/5598f8357944eeb6e5c1885adba444f3cc1f86a9/src/UnitfulAstro.jl#L55

That's a good point. I do not know why they need this unit since "angstrom" is so small to be astronomical. I do not know whether it would impact people who write

using UnitfulAstro: angstrom

if we merge this PR into Unitful.jl and ask UnitfulAstro.jl maintainers to remove it.

But it sounds weird to me that I need to install UnitfulAstro.jl in my field: dealing with atoms and molecules.

[giordano]

I do not know why they need this unit since "angstrom" is so small to be astronomical

Wavelengths 🙂

if we merge this PR into Unitful.jl and ask UnitfulAstro.jl maintainers to remove it.

I'm sure we'll find a way to not break users' code if angstrom is added here 😉

[singularitti]

I would suggest adding a line const Å = angstrom so that people who want to type it are able to do so but otherwise I agree this is a well-known and common unit that should probably go here.

Was it necessary to write Unitful.nm in the macro?

The reason I used Unitful.nm is that it will keep the fractional-structure when converting

julia> @unit angstrom "Å" Angstrom (1 // 10) * Unitful.nm false

julia> Unitful.register(Main)

julia> ustrip(u"cm", 1u"angstrom")
1//100000000

However, if I use UnitfulAstro.jl, it will be converted to floats:

julia> using UnitfulAstro

julia> ustrip(u"cm", 1u"angstrom")
1.0e-8

However, u"m" and u"cm" are related by fractionals:

julia> ustrip(u"m", 1u"cm")
1//100

So the relation breaks.
I do not know whether this is a good choice. But I saw you prefer fractional-structures:

@unit inch      "inch"     Inch                 (254//10000)*m          false
@unit mil       "mil"      Mil                  (1//1000)*inch          false

[singularitti]

I would suggest adding a line const Å = angstrom so that people who want to type it are able to do so but otherwise I agree this is a well-known and common unit that should probably go here.

Hi, would you mind pointing out where should I add const Å = angstrom?

[ajkeller34]

const Å = angstrom should go immediately following the @unit definition.

[ajkeller34]

The reason I used Unitful.nm is that it will keep the fractional-structure...

My question was really about whether the Unitful. part of Unitful.nm was necessary. Why not just write nm, consistent with the style of the rest of the file?

Note that Angstrom is already in UnitfulAstro.jl

At least the u_str macro shouldn't fail silently if you have two different modules defining the same unit—I think the unit from the most recently registered units module is used, and importantly a warning should be emitted.

[codecov-io]

Codecov Report

Merging #271 into master will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master     #271   +/-   ##
=======================================
  Coverage   78.34%   78.34%           
=======================================
  Files          15       15           
  Lines        1076     1076           
=======================================
  Hits          843      843           
  Misses        233      233

Impacted Files	Coverage Δ
src/pkgdefaults.jl	20.4% <ø> (ø)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 5bdf76e...12fe3af. Read the comment docs.

[singularitti]

The reason I used Unitful.nm is that it will keep the fractional-structure...

My question was really about whether the Unitful. part of Unitful.nm was necessary. Why not just write nm, consistent with the style of the rest of the file?

Note that Angstrom is already in UnitfulAstro.jl

At least the u_str macro shouldn't fail silently if you have two different modules defining the same unit—I think the unit from the most recently registered units module is used, and importantly a warning should be emitted.

Oh, I was just not aware of that. I have updated as you suggested.

[singularitti]

Any updated information?