# Eqnarray help (disadvantage) when breaking across columns

I'm a first time TeX user and had typed out certain (long) equations in a single column format for submission to a journal (that part worked fine). Unfortunately, I accidentally missed the fact (!) that Physical Review journals have a two column format. Directly converting to the two column format results in them spilling over the column width. So, now, I need to break all these beautiful long equations so that nothing spills over columns.

Q.1 - Is there any way to do this automatically, without changing too much (time constraint)?

Q.2 - Using eqnarray has one disadvantage -

\begin{eqnarray}

LHS & = & RHS1 \nonumber \\\

& & + RHS2 \nonumber\\\

& & + RHS 3

\end{eqnarray}


Of course I want the three RHS to be aligned, so that the equation displays well, but even if LHS is not a very big string, I'm wasting some space = width of LHS = '' (which is conspicuously large in a two column format).

(THOSE OF YOU WHO ARE PLANNING TO MARK THIS AS A DUPLICATE, PLEASE ENSURE THAT YOU REFER TO THE RIGHT POST SO THAT MY QUERY GETS ADEQUATELY ANSWERED. THANKS.)

PS - As demanded by @egreg and @daleif, I'm also adding a sobered up Minimum Working Example (MWE) -

\documentclass{article}

\begin{document}

\begin{eqnarray}
\Pi(\rho, \sigma, \delta) & = & \frac{1}{B} (4\rho_p + 2\rho_n + \rho_{\Lambda} + 2\rho_{\Sigma^+})-2\rho_{\Sigma^-} + 2\rho_{\Sigma^0} + 2\rho_{\Xi^0} - 2\rho_{\Xi^-}) - \frac{(\sigma_1 + (\sigma_3)_z)}{2\sigma_2\delta} \nonumber\\
& & + \frac{\sigma_1}{\sqrt{2}\delta}(4\rho^s_p + 2\rho^s_n + \rho^s_{\Lambda} + 2\rho^s_{\Sigma^+})-2\rho^s_{\Sigma^-} + 2\rho^s_{\Sigma^0} + 2\rho^s_{\Xi^0} - 2\rho^s_{\Xi^-})
\end{eqnarray}

\end{document}


Now, of course, this has been broken' at the appropriate places for a single column format, but needs to be revamped for a two column format.

• Welcome to TeX.SX! You can be interested in eqnarray vs align which gives you strong motivations to avoid eqnarray. BTW: your question seems to me a duplicate of How to break a long equation?. – Claudio Fiandrino May 12 '14 at 6:21
• Please help us to help you and add a minimal working example (MWE) that illustrates your problem. It will be much easier for us to reproduce your situation and find out what the issue is when we see compilable code, starting with \documentclass{...} and ending with \end{document}. – egreg May 12 '14 at 6:50
• I am providing a more sobered up MWE\\\documentclass{article} \begin{document} \begin{eqnarray} \Pi(\rho, \sigma, \delta) & = & \frac{1}{B} (4\rho_p + 2\rho_n + \rho_{\Lambda} + 2\rho_{\Sigma^+})-2\rho_{\Sigma^-} + 2\rho_{\Sigma^0} + 2\rho_{\Xi^0} - 2\rho_{\Xi^-}) - \frac{(\sigma_1 + (\sigma_3)_z)}{2\sigma_2\delta} \nonumber\\ & & + \frac{\sigma_1}{\sqrt{2}\delta}(4\rho^s_p + 2\rho^s_n + \rho^s_{\Lambda} + 2\rho^s_{\Sigma^+})-2\rho^s_{\Sigma^-} + 2\rho^s_{\Sigma^0} + 2\rho^s_{\Xi^0} - 2\rho^s_{\Xi^-}) \end{eqnarray} \end{document} – TeX beginner May 12 '14 at 7:19

Here is what I usually do using mathtools to pull the first line back

\documentclass{article}
\usepackage{mathtools}
\begin{document}

\begin{align}
\MoveEqLeft
\Pi(\rho, \sigma, \delta) =  \frac{1}{B} (4\rho_p + 2\rho_n +
\rho_{\Lambda} + 2\rho_{\Sigma^+})
\nonumber \\
& -2\rho_{\Sigma^-} +
2\rho_{\Sigma^0} + 2\rho_{\Xi^0} - 2\rho_{\Xi^-}) - \frac{(\sigma_1
+ (\sigma_3)_z)}{2\sigma_2\delta} \nonumber
\\
&  + \frac{\sigma_1}{\sqrt{2}\delta}(4\rho^s_p + 2\rho^s_n +
\rho^s_{\Lambda} + 2\rho^s_{\Sigma^+})
\nonumber \\
&-2\rho^s_{\Sigma^-} +
2\rho^s_{\Sigma^0} + 2\rho^s_{\Xi^0} - 2\rho^s_{\Xi^-})
\end{align}

\end{document}


Note that you have some unopened )'s.

• Thanks. Firstly, those unopened braces were a mistake. Secondly (and importantly), this way, I'm not following what advantage does align give over eqnarray. You are still breaking them manually. Does that mean there's no way to do it automatically? I mean whenever, it reaches the end of the column width, it automatically goes over to the next line, and not spill over, out of the column. – TeX beginner May 12 '14 at 9:56
• No, the breaks are always manual. Align is prefered because eqnarray contains a lot of errors and has a non consistent typography. breqn` can do automatic breaks, but has a lots of other issues. Most people break this manually, because then they are in charge of how it should look. It is very hard to write down explicit rules as to how math should be broken and aligned afterwards. – daleif May 12 '14 at 9:58