# scalebox scales the equation number in a multiline equation nested in subequations

I have a list of equations in a subequations environment. Some of the equations are too big, so I scale them down with scalebox. However the equation number gets scaled down with the rest. How can I prevent that?

I tried replacing gather by align but then I get an error. ("Erroneous nesting")

Here is a MWE:

\documentclass[a4paper,11pt]{article}

\usepackage{amsmath}
\usepackage{graphics}

\begin{document}

\begin{gather}
A+B = C \\
\scalebox{.7}{\begin{align} A+B+C & = D+E+F \notag \\ & \qquad +G+H+I \end{align}}
\end{gather}

\end{document}


• You should not scale equations (nor tables). This yields inconsistent font sizes and arduous reading. There are plenty of multiline environments you can use. Oct 26, 2020 at 11:26
• @Bernard I would agree with that statement most of the time, but I have a huge equation in an appendix that takes more than an entire page at normal font. Moreover, the terms are so long that I have to break them linewise into products. I think the equation is actually more readable in its scaled version.
– Pxx
Oct 26, 2020 at 11:30
• In this case, you can use one of the medium size commands and environment from nccmath. It's about 80 % of \displaystyle. Just out of curiosity, is it a formula like those in Invariant theory? Oct 26, 2020 at 11:39
• @Bernard I don't know about invariant theory, I do theoretical physics (conformal bootstrap to be precise), and there the functions (conformal blocks) are lengthy and not abbreviatable (they are already abbreviated as far as it goes and they are still huge!).
– Pxx
Oct 26, 2020 at 12:35
• It's a theory very much in favour during the 2nd half of the 19th century. I've see in a textbook of the time formulæ requiring up to 3pages! Oct 26, 2020 at 12:43

First of all: I agree with Bernard's comment that scaling an equation is usually a bad idea. In my experience of long, horrible equations the best solution has always been to introduce meaningful abbreviations.

This being said, you should let gather do the job of numbering the equation. The strange output you are getting is due to using align within inline math $...$. It is a funny coincidence that this does not raise an error (because you are in a gather environment). The solution is thus simply to replace align with its internal form aligned. In order to have the second line numbered use the b optional parameter (which aligns the baseline).

\documentclass[a4paper,11pt]{article}

\usepackage{amsmath}
\usepackage{graphicx}

\begin{document}

\begin{gather}
A+B = C \\
\scalebox{.7}{\begin{aligned}[b] A+B+C = D &+ E+F \\ &+ G+H+I \end{aligned}}
\end{gather}

\end{document}


The (horrible :-)) output:

• Thanks a lot for your answer! Of course, the output is horrible with the example I gave, but it looks fine with my actual equation!
– Pxx
Oct 26, 2020 at 12:38

As I mentioned it in a comment, here is how to do it with nccmath (which loads amsmath, but has to be loaded before mathtools if you need it):

\documentclass[a4paper,11pt]{article}

\usepackage{amsmath, nccmath}

\begin{document}

\begin{subequations}
\begin{gather}
A+B = C \\
\medmath{
\begin{aligned}[b]
A+B+C = D &+ E+F \\
&+ G+H+I
\end{aligned}}
\end{gather}
\end{subequations}

\end{document}