The Virtue of Numbering ALL your Equations

post by Linda Linsefors · 2017-09-28T18:41:35.631Z · LW · GW · 4 comments


  A is strictly better than B

Epistemic status: This is my strongly hold preference, but I dont know to what extent others agree.

I know of two common styles for numbering equations in a scientific publications.

A) Number ALL equations.

B) Number only those equations that you yourself want to reference to in the surrounding text.

A is strictly better than B

  • There is no extra effort to in numbering all equations since LaTeX does it for you.

  • It becomes much easier to discuss your paper in text and/or online, and in any other situation where the persons involved can’t just point directly at the equation they wants to refeer to.

LaTeX automatically number your equations if you use

\begin{equation} … \end{equation}

LaTeX does not automatically number your equation if you use

$$ … $$

Here is a bit of code you can use if you think “\begin{equation} … \end{equation}” Is too much to write. Copy paste this


into your tex document, before “\begin{document}”. Now you can simply use

\be … \ee

instead of “\begin{equation} … \end{equation}”

Please number all your equations!


Comments sorted by top scores.

comment by drossbucket · 2017-09-30T06:49:20.994Z · LW(p) · GW(p)

My second supervisor for my PhD was a big fan of this short essay by David Mermin, which you might like. He got all the new students to read it, and insisted on us always following the three rules there:

Rule 1: Number all your equations.

Rule 2: When referring back to an equation, identify it with a phrase as well as a number so the reader knows what you're talking about.

Rule 3: Punctuate equations like prose.

Good advice for helping the reader along.

comment by philh · 2017-09-29T10:07:57.178Z · LW(p) · GW(p)

It seems to me that numbering only some equations might be valuable for emphasis; "pay extra attention to this one". But equations you want to emphasise won't correlate perfectly with ones you want to refer to later, so it's not a great solution.

Do you happen to have thoughts on how you could emphasise equations without numbering them?

comment by Gunnar_Zarncke · 2017-09-29T07:40:12.854Z · LW(p) · GW(p)

And I thought it was about *all* equations, not just those in articles/papers...
But seriously it reminds me of the practice of a) dating all your notes (which I do) and b) numbering all your notes (which I sadly don't). The latter is the practice much lauded of Niklas Luhmann and his Zettelkasten. See Luhmanns Zettelkasten (English) and Wikipedia on Zettelkasten (Deutsch)