I am using the commath package to typeset derivatives and need to use a superscript in the argument supplied to \dpd. What I require is something like

    \dpd{}{X^{(j)}_0} f(X^{(j)}_0)

but this does not compile. There is no problem with subscripts as

    \dpd{}{X_0} f(X^{(j)}_0)

does compile. Does anyone know of a workaround?

(Note sure about the tag used - perhaps someone can suggest something more appropriate?)

  • I realise one may use $\frac{\partial}{\partial X^{(j)}_0} f(X^{(j)}_0)$ but I would like to know if it can be done within the \dpd environment. – Freakalien Feb 28 '16 at 3:13
  • I have not used commath package and think it was because there were some issues with it. I haven't found where I thought I read a recommendation to not use this package, but perhaps you should look at this question and the comments and answer: commath and \ifinner. Just an FYI. – Peter Grill Feb 28 '16 at 3:50
  • @Petter Grill I have noticed that the \pd (and \od) environment does a poor job of automatically determining whether a displayed or inline equation is required, but this can be overcome by specifying this manually as \tpd or \dpd etc. – Freakalien Feb 28 '16 at 4:06
  • really I wouldn't use commath, so many of its macros are incorrect. – David Carlisle Feb 28 '16 at 14:24
  • @David Carlisle yeah, I see your point. I wish I had used a macro to define the derivatives - a lot quicker to change than going step by step through the document. – Freakalien Feb 29 '16 at 4:13

Please observe additional parentheses in the solution:



    \dpd{}{{X^{(j)}_0}} f(X^{(j)}_0)


Nevertheless, this package seems to be abandoned. As we cannot expect correcting its bugs, I wouldn't recommend using its macros.

  • Do you know which is the best alternative? – Freakalien Feb 28 '16 at 4:10
  • @Freakalien It is easy to write your own macros, remembering about potential double superscript, e.g. \def\dpd#1#2{\dfrac{\mathrm{d}{#1}}{\mathrm{d}{#2}}}. – Przemysław Scherwentke Feb 28 '16 at 4:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.