6

I am using the commath package to typeset the derivative operator "D" with the command \Dif. This work well unless I want to add subscripts or superscripts. In this case, the spacing after operator is too small and the exponent / index "touches" the operator:

enter image description here

where the first operator is created via \Dif and the second one via \mathrm{D} (full MWE see below).

Is there some way to "redefine" the \Dif command in order to fix the spacing?

\documentclass{article}
\usepackage{commath}

\begin{document}

\begin{align*}
  \Dif^2_x f(x) \\
  \mathrm{D}^2_x f(x)
\end{align*}

\end{document}
Moriambar
  • 11,466
Bernd
  • 1,785
  • 1
    {\Dif\,}^2_x will do it, but that might be a nuisance, and not generally applicable. – Steven B. Segletes Jul 10 '13 at 12:02
  • Is there a way to make this automatically happen in general? I.e. to make \Dif work as it currently does almost always, but to produce something like {\Dif\,}^2_x when sub/superscripts are used? – Abe Schulte Apr 14 '14 at 21:20
  • @AbeSchulte What's the problem in using the definition I proposed in my answer? The definition by commath is inherently wrong and should never be used. – egreg Apr 14 '14 at 21:47
  • 1
    http://tex.stackexchange.com/questions/14821/whats-the-proper-way-to-typeset-a-differential-operator

    It seems there is quite a bit of disagreement regarding what exactly is "right" and therefore what is wrong. Unless you can cite a source, I really don't think you can say something is "inherently wrong", especially when an international organization entirely dedicated to stuff like this says otherwise (see the top answer of the linked post).

     – Abe Schulte Apr 15 '14 at 01:58
  • 1
    My suggestion: \renewcommand{\Dif}{\operatorname{D}} – Robert Fuster Apr 18 '14 at 11:19
  • @AbeSchulte I wasn't referring to whether the D should be upright or slanted, nor to any other typographic aspect; I'm saying that the definition of \Dif given in commath is TeXnically wrong and the only way for making it work is to modify it. – egreg Apr 20 '14 at 10:11
  • Could you explain why \DeclareMathOperator{\Dif}{D \!} is TeXnically wrong, and why \newcommand{\Dif}{\mathop{}\!D} is superior (or point me to a place that explains this)? – Abe Schulte Apr 21 '14 at 22:26

2 Answers2

11

The definition of \Dif in commath.sty is

\DeclareMathOperator{\Dif}{D \!}

which is simply wrong. The \! backspace is the cause of your problem. It's not the only part in commath that I consider wrong, so I can't recommend using the package.

A more correct definition would be

\newcommand{\Dif}{\mathop{}\!D}

that you can add to your document (with \renewcommand if you insist in using commath).

egreg
  • 1,121,712
  • 1
    Thanks a lot for this! Just out of curiosity - could you elaborate on why commath should not be used? If I exclude it, I better do so now (i.e. before relying on too many of its commands). – Bernd Jul 10 '13 at 12:09
  • @Bernd See, for instance, http://tex.stackexchange.com/a/120379/4427 – egreg Jul 10 '13 at 16:48
  • Ah, I see. I saw e.g. (http://tex.stackexchange.com/questions/22076/can-i-have-a-flexible-partial-derivative-macro) that the cool package seems to accomplish some things imilar to commath. Do you also recommend avoiding this package, or is it "kosher"? – Bernd Jul 11 '13 at 07:56
  • @Bernd I never used cool, so I don't know about it. – egreg Jul 11 '13 at 08:31
  • @egreg, what is your opinion about using the package esdiff for typesetting differentials? – mas Apr 15 '14 at 02:01
-1
\documentclass{article}

\usepackage{commath}

\begin{document}

\begin{align*}
\Dif^{^2}_{_x} f(x) \\
\mathrm{D}^{^2}_{_x} f(x)
\end{align*}

\end{document}

enter image description here

egreg
  • 1,121,712
Micky K
  • 1,457
  • 5
    Using double sub- and superscripts to avoid kerning problems is never a good solution. It makes your equation much more difficult to read because of the smaller symbols and the text will get inconsistent if you use “normal” sub- and superscripts in other equations (or other contexts in the same equation). – Lupino Apr 20 '14 at 09:30