# How to get only one vertically centered equation number in align environment with two equations

What I have are two equations with align and I want one single equation number to refer to this system of equations. To illustrate, I have

\begin{align}
dr_t &= \kappa ( \theta - r_t ) dt + \sigma_r r^{\xi} dW_t \\
dA_t &= \mu A_t dt + \sigma_A^{\alpha} dZ_t,
\end{align}


and I want neither

\begin{align}
dr_t &= \kappa ( \theta - r_t ) dt + \sigma_r r^{\xi} dW_t \nonumber \\
dA_t &= \mu A_t dt + \sigma_A^{\alpha} dZ_t,
\end{align}


nor

\begin{align}
dr_t &= \kappa ( \theta - r_t ) dt + \sigma_r r^{\xi} dW_t \\
dA_t &= \mu A_t dt + \sigma_A^{\alpha} dZ_t, \nonumber
\end{align}


but having a single number in the middle. I know this can be achieved by nesting an array inside of an equation but then some symbols (sums, etc.) look differently and I thought there had to be a better way of doing this.

Use aligned instead.

\documentclass{article}
\usepackage{amsmath}
\begin{document}
\begin{aligned} dr_t &= \kappa ( \theta - r_t ) dt + \sigma_r r^{\xi} dW_t \\ dA_t &= \mu A_t dt + \sigma_A^{\alpha} dZ_t, \end{aligned}
\end{document}


• @Paul: That seems wrong. Hmm. – TH. Mar 14 '11 at 15:11
• @Martin: I think it was rejected because most of the TeX code here is designed to display as TeX, not as graphics. – TH. Mar 14 '11 at 17:34
• also with the complicated LaTeX constructions sometimes shown here, it may be difficualt to secure a LaTeX to image renderer. Depending on the system it might e.g. be easy to fill a disk/partition or make LaTeX run in an infinite loop. Thus I thinkthe current situation is the best compromise. – daleif Mar 22 '11 at 23:05
• @TH.: I have a comment on your answer which is too big to give as a comment, so I've put it as a separate (noncompeting) answer. – Ryan Reich May 5 '11 at 15:05
• Environments equation and align serve similar kinds of purposes. Whereas the purpose of split or alignedis quite distinct: i.e. they typically fit inside equation, align*, etc. – Evgeni Sergeev Jul 12 '17 at 5:14

Here's a slightly better way to do this (IMO):

Using the split environment within an align will produce one equation number, vertically centered within the split (so long as there's not a page break involved, as discussed here).

Note that you can have individual labels for each equation by adding \label{} to the end of each line before the line break \\

For example:

\documentclass{article}
\usepackage{amsmath}
\usepackage{amssymb}

\begin{document}

\begin{align}
E_1&=A+B \label{eq:1}\\
\begin{split}
E_2&=(C-D)E_1 \label{eq:2}\\
\end{split}\\
E_3 &=(\pi\cdot \chi)-(R\cdot E_1)-(RY\delta\cdot E_2) \label{eq:3}
\end{align}

\end{document}


which gives:

This is really a comment on TH.'s accepted answer (which should stay that way). Using just one aligned doesn't work if you have multiple columns: they are separated by only a small space, since aligned "shrinks to fit". The workaround is to use one aligned per column, and wrap the whole thing in a larger align, like so:

\documentclass{article}
\usepackage{amsmath}
\begin{document}
\begin{align}
\begin{aligned}
x &= y \\       f(x) &= f(y)
\end{aligned}
&&
\begin{aligned}
a &= b \\       g(a) &= g(b)
\end{aligned}
\end{align}
\end{document}


This should produce two pairs of equations, each aligned at the equals sign, and with one equation number centered between the lines.

• It would seem better to wrap the whole thing in a larger equation, and separated the aligned by using a \quad or two. – Teepeemm Nov 13 '19 at 18:33
• I disagree; explicit positioning is exactly what we use aligned to avoid. If you want greater control (and I don't think this is visually critical enough to warrant designing your document code around this goal) you can play with the column spacing. – Ryan Reich Nov 15 '19 at 18:30
• One problem is that when the equations on different columns have different heights, the horizontal alignment gets screwed. – stafusa Feb 11 '20 at 14:09

Using

$$\begin{split} dr_t &= \kappa ( \theta - r_t ) dt + \sigma_r r^{\xi} dW_t \\ dA_t &= \mu A_t dt + \sigma_A^{\alpha} dZ_t, \end{split}$$


will produce one equation number that is vertically centered on the group of equations.