# Rescaling equation in align+split environment - Numbering problem

I'm trying to rescale my to long equation in align + split environment (to get only one numbering). I found a method from here which is pretty useful.

However, the numbering is altered and is indicated below the equations instead of at the right. I guess it's due to the \parbox command but I don't know how to fix it.

Here is a MWE, with a long equation:

\documentclass{article}

\usepackage{graphicx}
\usepackage{amsmath}

\begin{document}
\scalebox{0.8}{\parbox{.5\linewidth}{%
\begin{align}
\begin{split}
C_{M,i}(x) = &-\frac{e^{\text{Ha}_{i} - x \, \frac{k}{D_{\text{NO},t}}}\left(\left(-1 + e^{\text{H\~{a}}_{i}}\right)\left(-1+\text{Mu}_{i}+e^{\text{H\~{a}}_{i}}(1+\text{Mu}_{i})\right) \, Pr + 2 \, e^{\text{H\~{a}}_{i}} \, k \, \lambda_{\text{air},i} \, C_{i}(t,z) \right)}{2 \, k \left(-1 + \text{Mu}_{i} + e^{2 \, (\text{Ha}_{i}+\text{H\~{a}}_{i})}(1+\text{Mu}_{i})\right)} \\
&+\frac{e^{\text{Ha}_{i} + x \, \frac{k}{D_{\text{NO},t}}}\left(\left(-1+e^{\text{H\~{a}}_{i}}\right)\left(-1+\text{Mu}_{i} + e^{\text{H\~{a}}_{i}}(1+\text{Mu}_{i})\right) \, Pr + 2 \, e^{\text{H\~{a}}_{i}} \, k \, \lambda_{\text{air},i} \, C_{i}(t,z)\right)}{2 \, k \, \left(-1+\text{Mu}_{i} + e^{2(\text{Ha}_{i} + \text{H\~{a}}_{i})}(1+\text{Mu}_{i})\right)}
\end{split}
\end{align}
}}
\end{document}

• It's never a good idea to scale equations but if you do it this way you are scaling after the number is added so if the equation number is moved because of the long equation, the scaling will not change that it just makes it bigger. But please fix the example so it is a complete example that people can run to see the problem. May 24 '17 at 10:21
• @DavidCarlisle I corrected the example with a long enough equation May 24 '17 at 11:21
• I made it a proper example that shows the problem, but as I said in the first comment the \scalebox isn't changing the positioning of the equation number. May 24 '17 at 11:36

If you really want to scale the equations, do it properly:

\documentclass{article}
\usepackage{amsmath,graphicx}

\begin{document}

\resizebox{.8\textwidth}{!}{% \begin{aligned} C_{M,i}(x) ={} & -\frac{ e^{\mathrm{Ha}_{i} - x \frac{k}{D_{\mathrm{NO},t}}} \bigl( (-1 + e^{\mathrm{H\tilde{a}}_{i}})(-1+\mathrm{Mu}_{i} + e^{\mathrm{H\tilde{a}}_{i}}(1+\mathrm{Mu}_{i}))Pr + 2e^{\mathrm{H\tilde{a}}_{i}} k\lambda_{\mathrm{air},i} C_{i}(t,z) \bigr) }{ 2k(-1 + \mathrm{Mu}_{i} + e^{2(\mathrm{Ha}_{i}+\mathrm{H\tilde{a}}_{i})}(1+\mathrm{Mu}_{i})) } \\ & +\frac{ e^{\mathrm{Ha}_{i} + x \frac{k}{D_{\mathrm{NO},t}}} \bigl( (-1+e^{\mathrm{H\tilde{a}}_{i}})(-1+\mathrm{Mu}_{i} + e^{\mathrm{H\tilde{a}}_{i}}(1+\mathrm{Mu}_{i}))Pr + 2e^{\mathrm{H\tilde{a}}_{i}} k\lambda_{\mathrm{air},i} C_{i}(t,z)) }{ 2k(-1+\mathrm{Mu}_{i} + e^{2(\mathrm{Ha}_{i} + \mathrm{H\tilde{a}}_{i})}(1+\mathrm{Mu}_{i})) } \end{aligned}% }

\end{document}


I removed all \left and \right; also all \, commands (they were wrong). I only used \bigl( and \bigr) in the outermost parentheses of the big expressions.

Don't use \text, but \mathrm in this context.

I think you (and even more the readers of your document) would be better off if you didn't scale down the equation to make it fit inside the text block. Instead, consider using the \splitdfrac macro, which is provided by the mathtools package, to split the long numerator terms into two parts.

\documentclass{article}
\usepackage{mathtools} % for 'splitdfrac'  environment
\begin{document}

\begin{aligned}[b] C_{M,i}(x) &= -\frac{\left(\splitdfrac{e^{\textrm{Ha}_i - x \frac{k}{D_{\textrm{NO},t}}} \bigl[(-1+e^{\textrm{H\~{a}}_i})(-1+\textrm{Mu}_i}{+e^{\textrm{H\~{a}}_i}(1+\textrm{Mu}_i)) Pr + 2e^{\textrm{H\~{a}}_i} k \lambda_{\textrm{air},i} C_i(t,z) \bigr]}\right)}{ \displaystyle 2k \bigl[ -1+\textrm{Mu}_i + e^{2(\textrm{Ha}_i+\textrm{H\~{a}}_i)} (1+\textrm{Mu}_i) \bigr]} \\[2ex] &\quad +\frac{\left(\splitdfrac{e^{\textrm{Ha}_i + x \frac{k}{D_{\textrm{NO},t}}} \bigl[(-1+e^{\textrm{H\~{a}}_i})(-1+\textrm{Mu}_i}{+e^{\textrm{H\~{a}}_i}(1+\textrm{Mu}_i)) Pr + 2e^{\textrm{H\~{a}}_i} k \lambda_{\textrm{air},i} C_i(t,z) \bigr]}\right)}{ \displaystyle 2k \bigl[ -1+\textrm{Mu}_i + e^{2(\textrm{Ha}_i+\textrm{H\~{a}}_i)} (1+\textrm{Mu}_i) \bigr]} \end{aligned}
\end{document}