[Tex/LaTex] Break line within array or eqnarray

arrayseqnarraymultline

I have used the following codes to generate two equations. However, the second equation overlaps with the equation index. Can someone please help me look it pretty by breaking the 2nd equation into two lines? By the way, I also read on stackexchange to avoid eqnarray and instead use "array". Could you help me with that too? Apparently, using these codes within array resulted in error.

\begin{eqnarray}
\lambda_{R,t} &=& \dfrac{(C_{R,t}-\phi_c C_{R,t-1})^{-\sigma}} 
{P_t(1+\tau_t^c)} - \phi_c \beta \dfrac{(E_t C_{R,t+1}-\phi_c C_{R,t})^{- 
\sigma}}{P_t(1+\tau_t^c)} \\
Q_t &=& \beta E_t{(1-\delta)Q_{t+1} + \lambda_{R,t+1}R_{t+1}U_{t+1}(1- 
\tau^k_{t+1})} - \lambda_{R,t+1}P_{t+1} \left[\Psi_1(U_t-1) + \dfrac{\Psi_2} 
{2}(U_t-1)^2 \right]
\end{eqnarray}

Best Answer

The environment align is probably the way to go here:

\documentclass[12pt]{article}
\usepackage{amsmath}

\begin{document}


\begin{align}
\lambda_{R,t} &= \dfrac{(C_{R,t}-\phi_c C_{R,t-1})^{-\sigma}} 
{P_t(1+\tau_t^c)} - \phi_c \beta \dfrac{(E_t C_{R,t+1}-\phi_c C_{R,t})^{- 
\sigma}}{P_t(1+\tau_t^c)} \\
Q_t &= \beta E_t{(1-\delta)Q_{t+1} + \lambda_{R,t+1}R_{t+1}U_{t+1}(1- 
\tau^k_{t+1})} \\ 
& - \lambda_{R,t+1}P_{t+1} \left[\Psi_1(U_t-1) + \dfrac{\Psi_2} 
{2}(U_t-1)^2 \right] \nonumber
\end{align}


\end{document}

You also have to use \nonumber in order to avoid having three equation numbers.

Hope that helps.

Romain

Related Solutions

[Tex/LaTex] Numbering an array as one single equation LaTeX

Using the eqnarray environment instead of array, you can add \nonumber to any lines you don't want numbered. If you don't do this, every line will be numbered.

\begin{eqnarray}
N_{x,sk} &=& k_{sk}\left(\frac{t_{sk}}{b_{sk}}\right)^{2}\bar{Et} \nonumber \\
N_{x,st} &=& k_{st}\left(\frac{t_{st}}{b_{st}}\right)^{2}\bar{Et}
\end{eqnarray}

(Just remove the \nonumber from the above to get the second example.)

[Tex/LaTex] Line break within an align-environment

IMHO, splitting long equations should always be done with more care from the author. So I always split my equations manually to get the desired output.

My preferred solution below uses the following tricks:

A \phantomrel command which creates a “phantom” relation symbol that occupies the same space as the underlying symbol, thus the \phantomrel{=}.
A \hphantom{\int_0^T \int_{K_1}} that creates a horizontal phantom space which serves as indent for the \times ... contents.
A \diff command dedicated for differentials.
A \abs command dedicated for absolute values, which must be created using \DeclarePairedDelimiter from the mathtools package. This \abs{...} has several advantages over |...|:
- Try typing $|-x|$ and $\abs{-x}$ to see the wrong spacing of the negative sign in the former expression.
- You get to enlarge the delimiters by variants \abs[\big]{-x}, \abs[\Big]{-x}, \abs[\bigg]{-x} and \abs[\Bigg]{-x}.
- There is another variant \abs*{-x} which extends delimiters automatically (like a \left and \right pair), but acts like \mathopen...\mathclose and not like \mathinner, which is better.
Please don’t abuse \left and \right. Sometimes they create delimiters that are too big or too small, and they create \mathinner. I used manual sizing throughout your example.

The environments you mentioned work perfectly fine.

\documentclass{article}
\usepackage{mathtools}% For `multlined' and loads `amsmath'
\newcommand*\phantomrel[1]{\mathrel{\phantom{#1}}}% My preferred typesetting
\newcommand*\diff{\mathop{}\!d}% Just in case your editor asks for \mathrm{d}
\DeclarePairedDelimiter\abs{\lvert}{\rvert}

\begin{document}
\noindent
Using \verb|align*| and \verb|multlined| works fine:
\begin{align*}
A & = \text{something something something} \\
  & = \begin{multlined}[t][10cm]
        \text{something something something anything anything} \\
        + \text{anything anything anything}
      \end{multlined} \\
  & = \text{anything anything anything anything}
\end{align*}
However, I~would typeset as follows instead:
\begin{align*}
A & = \text{something something something} \\
  & = \text{something something something anything} \\
  & \phantomrel{=} {} + \text{anything anything anything anything} \\
  & = \text{anything anything anything anything}
\end{align*}
So your specific example becomes
\begin{align*}
\mathcal{A} & = \int_0^T \int_{K_1}
    \Bigl[ c(\alpha,n) + \alpha\abs{u}^2
           - \alpha^2 \bigl(u\cdot(x-x_0)\bigr)^2
             \bigl(\abs{x-x_0}^2 + h^2\bigr)^{-1} \Bigr] \\
 & \phantomrel{=} \hphantom{\int_0^T \int_{K_1}}
    \times \bigl[\abs{x-x_0}^2 + h^2 \bigr]^\alpha \tau \diff x\diff t \\
 & \leq K \int_0^T \int_{K_1} \pi \diff x\diff t
\end{align*}
\end{document}

Best Answer

Related Solutions

[Tex/LaTex] Numbering an array as one single equation LaTeX

[Tex/LaTex] Line break within an align-environment

Related Question