Problem using alignat with 2 &&'s

Question

I am having trouble aligning the second block in my example below...

\documentclass{article}
\usepackage{amsmath}
\usepackage{amssymb}
\begin{document}
\begin{alignat}{2}\nonumber
    \frac{\partial U^\varepsilon_{p,q}}{\partial x_3} &= \frac{i\omega\Gamma_1(\kappa_1) m(x_3/\varepsilon^2)}{\varepsilon}\bigg(pU^\varepsilon_{p,q} - qU^\varepsilon_{p,q}\bigg) \ \nonumber
    &+\frac{i\omega\Gamma_2(\kappa_1) m(x_3/\varepsilon^2)}{2\varepsilon}e^{2i\omega\bar\zeta(\kappa_1)x_3/\varepsilon}&&\bigg(pU^\varepsilon_{p+1,q}e^{ih\bar\zeta(\kappa_1)x_3 - i\omega\bar{c}{66}\kappa_1\lambda x_3/(\bar{c}{44}\bar{\zeta}(\kappa_1))} \ \nonumber
    &\phantom{+} &&-qU^\varepsilon_{p,q-1}e^{-ih\bar\zeta(\kappa_1)x_3 + i\omega\bar{c}{66}\kappa_1\lambda x_3/(\bar{c}{44}\bar{\zeta}(\kappa_1))}\bigg) \ \nonumber
    &+\frac{i\omega\Gamma_2(\kappa_1) m(x_3/\varepsilon^2)}{2\varepsilon}e^{-2i\omega\bar\zeta(\kappa_1)x_3/\varepsilon}&&\bigg(pU^\varepsilon_{p-1,q}e^{-ih\bar{\zeta}x_3 + i\omega\bar{c}{66}\kappa_1\lambda x_3/(\bar{c}{44}\bar{\zeta}(\kappa_1))} \
    &\phantom{+} &&-qU^\varepsilon_{p,q+1}e^{ih\bar{\zeta}(\kappa_1)x_3 - i\omega\bar{c}{66}\kappa_1\lambda x_3/(\bar{c}{44}\bar{\zeta}(\kappa_1))}\bigg) 
\end{alignat}
\end{document}

The second line where \bigg(p appears should not have a large space between there and the exponential previous to it, could someone lend me a piece of advice please? Thanks.

You're using the wrong method. See math mode - How can I split an equation over two (or more) lines - TeX - LaTeX Stack Exchange — user202729, Nov 08 '21 at 15:35

score 2 · Answer 1 · answered Nov 08 '21 at 15:52

You might add a negative space in the first row to make TeX think it's shorter.

There will still be a small gap, due to the minus sign in the exponent, but I feel that alignment is more important.

\documentclass{article}
\usepackage{geometry}
\usepackage{amsmath}
\usepackage{amssymb}
\begin{document}
\begin{alignat}{2}
\frac{\partial U^\varepsilon_{p,q}}{\partial x_3}
&= \frac{i\omega\Gamma_1(\kappa_1) m(x_3/\varepsilon^2)}{\varepsilon}
   \bigl(pU^\varepsilon_{p,q} - qU^\varepsilon_{p,q}\bigr) \hspace{-5em}
\nonumber \
&+ \frac{i\omega\Gamma_2(\kappa_1) m(x_3/\varepsilon^2)}{2\varepsilon}
   e^{2i\omega\bar\zeta(\kappa_1)x_3/\varepsilon}
   &&\Bigl(pU^\varepsilon_{p+1,q}e^{ih\bar\zeta(\kappa_1)x_3 -
     i\omega\bar{c}{66}\kappa_1\lambda x_3/(\bar{c}{44}\bar{\zeta}(\kappa_1))}
\nonumber \
&  &&-qU^\varepsilon_{p,q-1}e^{-ih\bar\zeta(\kappa_1)x_3 +
     i\omega\bar{c}{66}\kappa_1\lambda x_3/(\bar{c}{44}\bar{\zeta}(\kappa_1))}\Bigr)
\nonumber \
&+ \frac{i\omega\Gamma_2(\kappa_1) m(x_3/\varepsilon^2)}{2\varepsilon}
   e^{-2i\omega\bar\zeta(\kappa_1)x_3/\varepsilon}
   &&\Bigl(pU^\varepsilon_{p-1,q}e^{-ih\bar{\zeta}x_3 +
     i\omega\bar{c}{66}\kappa_1\lambda x_3/(\bar{c}{44}\bar{\zeta}(\kappa_1))}
\nonumber \
&  &&-qU^\varepsilon_{p,q+1}e^{ih\bar{\zeta}(\kappa_1)x_3 -
     i\omega\bar{c}{66}\kappa_1\lambda x_3/(\bar{c}{44}\bar{\zeta}(\kappa_1))}\Bigr)
\end{alignat}
\end{document}

Note the usage of \Bigl and \Bigr for two reasons:

\bigg is too big;
one should set the delimiters to opening and closing atoms.

In the first line \bigl and \bigr seems more than enough.

I used geometry because, according to your picture, you have a larger text width than the default.

Bernard · Answer 2 · 2021-11-08T20:28:04.797

I propose this layout based on \multlined nested in aligned, using smaller parentheses, and the geometry package, to have more decent margins:

\documentclass{article}
\usepackage{geometry}
\usepackage{mathtools}
\usepackage{amssymb}
\begin{document}
\begin{equation}
\begin{aligned}[b]
    \frac{\partial U^\varepsilon_{p,q}}{\partial x_3} &=\frac{i\omega\Gamma_1(\kappa_1) m(x_3/\varepsilon^2)}{\varepsilon}\mathrlap{\Bigl(pU^\varepsilon_{p,q} - qU^\varepsilon_{p,q}\Bigr)} \
    &\begin{multlined} + \frac{i\omega\Gamma_2(\kappa_1) m(x_3/\varepsilon^2)}{2\varepsilon}e^{2i\omega\bar\zeta(\kappa_1)x_3/\varepsilon}\Bigl(pU^\varepsilon_{p+1,q}e^{ih\bar\zeta(\kappa_1)x_3 - i\omega\bar{c}{66}\kappa_1\lambda x_3/(\bar{c}{44}\bar{\zeta}(\kappa_1))} \[-1.5ex]
   -qU^\varepsilon_{p,q-1}e^{-ih\bar\zeta(\kappa_1)x_3 + i\omega\bar{c}{66}\kappa_1\lambda x_3/(\bar{c}{44}\bar{\zeta}(\kappa_1))}\Bigr)
    \end{multlined}\
    &\begin{multlined}[b] + \frac{i\omega\Gamma_2(\kappa_1) m(x_3/\varepsilon^2)}{2\varepsilon}e^{-2i\omega\bar\zeta(\kappa_1)x_3/\varepsilon}\Bigl(pU^\varepsilon_{p-1,q}e^{-ih\bar{\zeta}x_3 + i\omega\bar{c}{66}\kappa_1\lambda x_3/(\bar{c}{44}\bar{\zeta}(\kappa_1))} \[-1.5ex]
    -qU^\varepsilon_{p,q+1}e^{ih\bar{\zeta}(\kappa_1)x_3 - i\omega\bar{c}{66}\kappa_1\lambda x_3/(\bar{c}{44}\bar{\zeta}(\kappa_1))}\Bigr)
    \end{multlined}
\end{aligned}
\end{equation}
\end{document}

Problem using alignat with 2 &&'s

2 Answers2