How is a state of subscript notation in Julia 1.x?

I experiments with Julia subscript notation with 1.x version of language. It seems that is 30 or so available subscripts like A_1, A_2, A_x, A_schwa. I have impression that in versions 0.x this work different, but I can’t find these examples now.

Can someone explain to me how this looks now? I make small presentation about Julia and probably have time to find it myself.

I don’t think this changed. All the subscripts which are in unicode are defined, which are not all though:


I try to make A_y subscript in IJulia notebook and that don’t work, even as in Unicode table is defined. But I both learning Julia 1.x and have new operating system, which works quite strange, so I don’t know if problem is in Julia 1.x rules, IJulia or something with OS.

1 Like

Unfortunately, it’s one of the missing subscripts in unicode — see the table on the Wikipedia page @mauro3 linked. Unicode simply has not defined superscript q or subscripts b, c, d, f, g, q, w, y, z. It’s all very silly.

See GitHub - stevengj/subsuper-proposal: Draft proposal for additional sub/superscript characters in Unicode for a draft proposal to the Unicode standards committee to try to rectify the situation.


Some subscripts do not work in macOS either. I would really like the _k but I see an exclamation mark here.

1 Like

You just need to install a better font. I think Unifont is what ends up displaying many of these glyphs on my Mac (as a fallback, I think).

Thank you for pointing this out to me. I several times read “subscript” not “superscript” in this table and superscript A^y is defined. That is double silly, why subscripts are not compatible with superscripts?

As I understand it, each and every sup-/sub-script was independently argued-for and added in separate blocks for separate purposes. E.g., some might be for use with chemistry notations while others could be for phonetics. That’s why fonts don’t put them all at the same height or size, too.

1 Like

This is all because of whatever the Unicode Consortium decides. They have decided that “English letters” does not constitute a meaningful set of letters, so you have missing letters like q.

Please help with to try to get this fixed in Unicode.


I will do what I can.

1 Like

An interesting, but perhaps ambitious, project would be to mine relevant subsets (physics, stats, ML, optimization etc.) of arXiv for statistics on usage of sub/superscript in scientific articles.
On the flip side, github could be mined for variables named xx_i, x_2 etc.
Could such statistics perhaps be useful to make the arguments for the proposal?

Probably could. However, just polling a few field experts might be a quicker and safer way.

1 Like

Main problem to me is: what we can do to make basic update to Unicode? Already proposed update will be good thing at least. Maybe not perfect for many application, but definitely good thing.

  1. Ask nicely. The best way to do that is a proposal like the above. This may be slow, but cheap.

  2. Acquire enough shares in members of the Unicode Consortium enough for a majority vote. This may be faster, but rather expensive (of course we can sell them later and :partying_face:).

1 Like

I will follow first point. I start next week, hopefully.