6

In answering this question, I recommended some changes to the OP and provided the following code:

\documentclass[12pt]{article}
\pagestyle{plain}
\usepackage[margin=1.8cm]{geometry}
\geometry{a4paper}
\usepackage[parfill]{parskip}
\usepackage{amsmath}
\usepackage{amssymb}

\begin{document}

\[
  \begin{aligned}
    |f_{n}&(x)g_{n}(x) - f_{n}(x)g(x) + f_{n}(x)g(x) - f(x)g(x)| \\
    &\leq |f_{n}(x)g_{n}(x) - f_{n}(x)g(x)| + |f_{n}(x)g(x) -
    f(x)g(x)| \\
    &= |f_{n}(x)||g_{n}(x) - g(x)| + |g(x)||f_{n}(x) - f(x)| \\
    &\leq M_{1}\epsilon + M_{2}\epsilon \\
    &= \epsilon(M_1+M_2) \longrightarrow 0 \text{ as } n \to \infty
  \end{aligned}      
\]

\end{document}

Which produces

enter image description here

What I really wanted to do, though, is follow the advice of 3.3.5c of Mathematics Into Type, which recommends breaking at conjunctions and aligning with two-em quad from the left.

That I think is possibly open to interpretation, but what I would like is to have:

|f_{n}&(x)g_{n}(x) - f_{n}(x)g(x) + f_{n}(x)g(x) - f(x)g(x)|

On one line, a line break and all subsequent lines indented by a two-em quad and aligned. In other words, I want what I have, but with a precisely two-em quad indentation:

enter image description here

Unfortunately, the above is the closest I've been able to come. I tried:

\begin{align*}
  |f_{n}(x)g_{n}(x) - f_{n}(x)g(x) + f_{n}(x)g(x) - f(x)g(x)| \\
  \qquad &\leq |f_{n}(x)g_{n}(x) - f_{n}(x)g(x)| + |f_{n}(x)g(x) -
  f(x)g(x)| \\
  &= |f_{n}(x)||g_{n}(x) - g(x)| + |g(x)||f_{n}(x) - f(x)| \\
  &\leq M_{1}\epsilon + M_{2}\epsilon \\
  &= \epsilon(M_1+M_2) \longrightarrow 0 \text{ as } n \to \infty
\end{align*}

But that doesn't work at all:

enter image description here

4
  • Do you want the (in)equality signs below (x)?
    – Sigur
    Commented Feb 18, 2016 at 1:14
  • @Sigur Ah, sorry if I haven't explained myself properly. What I want is the (in)equality signs aligned with each other and indented exactly two-em quad from the start of the equation. Any equation, in fact. My hack - the best I could do - puts the signs more or less below (x) but what I'm aiming for is precisely two-em quad from the leftmost end of the equation. This was just an example. It looks alright this time, but in the next inequality, I may not be able to find a good spot to align and it may not be in the same place. I'm after a way to make the indentation two-em quad every time.
    – Au101
    Commented Feb 18, 2016 at 1:19
  • See tex.stackexchange.com/questions/250489/….
    – Zarko
    Commented Feb 18, 2016 at 1:29
  • 1
    He he, that exact picture (with the two lines) was exactly why \MoveEqLeft was added to mathtools as described below. I got a version of the image from Mathematics into type published by the AMS.
    – daleif
    Commented Feb 18, 2016 at 12:02

3 Answers 3

9

The mathtools provides the \MoveEqLeft function which achieves exactly what you want. By default, it indents subsequent lines by 2em and it can be further customized with \MoveEqLeft[<number>] which will indent subsequent line by <number> ems:

\documentclass{article}
\usepackage{amsmath}
\usepackage{mathtools}

\begin{document}
\begin{align}
  \MoveEqLeft |f_{n}(x)g_{n}(x) - f_{n}(x)g(x) + f_{n}(x)g(x) - f(x)g(x)| \\
  &\leq |f_{n}(x)g_{n}(x) - f_{n}(x)g(x)| + |f_{n}(x)g(x) - f(x)g(x)| \\
  &= |f_{n}(x)||g_{n}(x) - g(x)| + |g(x)||f_{n}(x) - f(x)| \\
  &\leq M_{1}\epsilon + M_{2}\epsilon \\
  &= \epsilon(M_1+M_2) \longrightarrow 0 \text{ as } n \to \infty
\end{align}      
\begin{align}
  \MoveEqLeft[4] |f_{n}(x)g_{n}(x) - f_{n}(x)g(x) + f_{n}(x)g(x) - f(x)g(x)| \\
  &\leq |f_{n}(x)g_{n}(x) - f_{n}(x)g(x)| + |f_{n}(x)g(x) - f(x)g(x)| \\
  &= |f_{n}(x)||g_{n}(x) - g(x)| + |g(x)||f_{n}(x) - f(x)| \\
  &\leq M_{1}\epsilon + M_{2}\epsilon \\
  &= \epsilon(M_1+M_2) \longrightarrow 0 \text{ as } n \to \infty
\end{align}      
\end{document}

output

2
  • The mathtools package loads amsmath automatically -- thus, no need to load amsmath explicitly.
    – Mico
    Commented Feb 18, 2016 at 10:43
  • @Mico I'm aware that mathtools loads amsmath, but I still prefer to load amsmath because a lot of IDEs read the usepackage declarations in order to load the appropriate commands and environments.
    – JP-Ellis
    Commented Feb 19, 2016 at 0:28
5

Is this what you want?

\begin{align*}
  &|f_{n}(x)g_{n}(x) - f_{n}(x)g(x) + f_{n}(x)g(x) - f(x)g(x)| \\
  &\qquad \leq |f_{n}(x)g_{n}(x) - f_{n}(x)g(x)| + |f_{n}(x)g(x) - f(x)g(x)| \\
  &\qquad = |f_{n}(x)||g_{n}(x) - g(x)| + |g(x)||f_{n}(x) - f(x)| \\
  &\qquad \leq M_{1}\epsilon + M_{2}\epsilon \\
  &\qquad = \epsilon(M_1+M_2) \longrightarrow 0 \text{ as } n \to \infty
\end{align*}

enter image description here

3
  • Yes! That's exactly the right output, thank you :) Obviously this approach has the downside of an inelegant input, but that's certainly what I wanted to produce :)
    – Au101
    Commented Feb 18, 2016 at 1:24
  • @Au101, see the elegant answer by JP-Ellis.
    – Sigur
    Commented Feb 18, 2016 at 1:25
  • 1
    Yes indeed, he will most likely get the tick, but I wanted to thank you for bearing with me and let you know you'd definitely understood my question :)
    – Au101
    Commented Feb 18, 2016 at 1:26
2

If you use mathenv of the mdwtools collection, you can use the enhanced {eqnarray} environment. It takes optional column specifiers:

  • r, c, l for right-justified, centered, and left-justified math;
  • L for left-justified math that is considered to have width 2em;
  • and more (read the documentation) so you can completely emulate the functionality of amsmath environments like {align} and others.

Here, you'd use

\documentclass{article}
\usepackage{amstext}
\usepackage{mathenv}

\begin{document}
\begin{eqnarray*}[Ll]
|f_{n}(x)g_{n}(x) - f_{n}(x)g(x) + f_{n}(x)g(x) - f(x)g(x)| \\
&\leq |f_{n}(x)g_{n}(x) - f_{n}(x)g(x)| + |f_{n}(x)g(x) -
f(x)g(x)| \\
&= |f_{n}(x)||g_{n}(x) - g(x)| + |g(x)||f_{n}(x) - f(x)| \\
&\leq M_{1}\epsilon + M_{2}\epsilon \\
&= \epsilon(M_1+M_2) \longrightarrow 0 \text{ as } n \to \infty
\end{eqnarray*}
\end{document}

mathenv eqnarray L demo

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .