1

I would like to make a command that produces the following results

\myquantity[Q]
\myquantity[Q][i]
\myquantity[Q][][j]
\myquantity[Q][i][j]
\myquantity[Q][i][j][\bar]
\myquantity[Q][][][\bar]
\myquantity[Q][][][\tilde]
\myquantity[Q_c][i][j][\bar]

Result

My current solution is the following

\newcommandx{\myquantity}[4][1=, 2=, 3=, 4=]{
  \des[#4]{#1}{#3}(\lambda_{#2})}

where \des is the command in the supplement of this answer.

Is there a simpler way to obtain the same result?

Simone Gaiarin
  • 217

2 Answers2

6

I would like to argue that having four optional arguments is not very semantic. I would use a command with a key=value optional argument.

\myquantity[Q_c][i][j][\bar]

is a lot less readable than

\myquantity[accent=\bar,subscript={c,j},lambda subscript=i]{Q}

although the latter might be using a bit too long keys, I hope you agree it is easier to understand.

The following MWE gives an example of your desired command, using a key-value argument. If you indeed think the keys are too long, you can always change them what you find descriptive enough. 

\documentclass{article}

\usepackage{amsmath}
\usepackage{pgf}

\makeatletter
\newif\if@myquantity@subscript
\newif\if@myquantity@lambdasubscript
\pgfkeys{
    my quantity/.cd,
        accent/.code={\let\@myquantity@accent#1},
        accent=\relax,
        subscript/.code={\def\@myquantity@subscript{#1}\@myquantity@subscripttrue},
        lambda subscript/.code={\def\@myquantity@lambdasubscript{#1}\@myquantity@lambdasubscripttrue},
}
\newcommand{\myquantity}[2][]{%
    \@myquantity@subscriptfalse
    \@myquantity@lambdasubscriptfalse
    \pgfkeys{my quantity/.cd,#1}%
    \if@myquantity@subscript%
        \@myquantity@accent{#2}_{\@myquantity@subscript}%
    \else%
        \@myquantity@accent{\mathbf{#2}}%
    \fi%
    (\lambda%
    \if@myquantity@lambdasubscript
        _{\@myquantity@lambdasubscript}%
    \fi)%
}
\makeatother

\begin{document}
    \begin{equation*}
        \begin{array}{lll}
            \myquantity{Q} &
            \myquantity[lambda subscript=i]{Q} &
            \myquantity[subscript=j]{Q}\\
            \myquantity[subscript=j,lambda subscript=i]{Q} &
            \myquantity[accent=\bar,subscript=j,lambda subscript=i]{Q} &
            \myquantity[accent=\bar]{Q} \\
            \myquantity[accent=\tilde]{Q} &
            \myquantity[accent=\bar,subscript={c,j},lambda subscript=i]{Q}
        \end{array}
    \end{equation*}
\end{document}

Resulting in

enter image description here

Max
  • 9,733
  • 3
  • 28
  • 35
  • Thanks for the answer. Where can I find a reference of the commands used in your answer? – Simone Gaiarin Aug 05 '18 at 14:29
  • @SimoneGaiarin the keys can be defined with the pgfkeys package which is automatically loaded when you're using Tikz or pgf. You can find more information on the key management in chapter 82 (page 875) of the Tikz/PGF manual . – Max Aug 05 '18 at 14:41
4

Four optional arguments require too much care in input. Besides, one of them is mandatory and the last one refers to the mandatory argument, so it should be specified first.

My idea is

\myq[<optional decoration>]{<variable>}[<optional coordinate>](<optional subscript)

so there is no need to specify empty optional arguments.

What do the macros do? First I look for _ in the optional argument, because this should not be part of \mathbf when no coordinate appears. For this purpose, \regex_split:nnN is available: if no _ is found, the sequence will have just one item, otherwise it will have two (what comes before _ and what comes after).

Then it's a matter of checking whether the coordinate is specified, in order to decide for \mathbf or not.

The <decoration> should be a math accent command; it defaults to \use:n, which means “output the argument” (the classical \@firstofone).

\documentclass{article}
\usepackage[leqno,fleqn]{amsmath} % the options are just to better show the output
\usepackage{xparse}

\ExplSyntaxOn

\NewDocumentCommand{\myq}{
  O{\use:n} % decoration for the main variable
  m         % main variable
  o         % coordinate
  d()       % subscript to \lambda
 }
 {
  % separate off the possible subscript to the variable;
  % item 1 contains the variable, item 2 the subscript
  \regex_split:nnN { \cD_ } { #2 } \l_simone_myq_var_seq
  % make the subscript
  \int_compare:nTF { \seq_count:N \l_simone_myq_var_seq = 1 }
   {% no subscript in #2
    \IfNoValueTF{#3}
     { #1 { \mathbf{#2} } }
     { #1 { #2 } \sb { #3 } }
   }
   {% subscript in #2
    \IfNoValueTF{#3}
     {
      #1 { \mathbf{\seq_item:Nn \l_simone_myq_var_seq { 1 }} } % base
      \sb { \seq_item:Nn \l_simone_myq_var_seq { 2 } }
     }
     {
      #1 { \seq_item:Nn \l_simone_myq_var_seq { 1 } } % base
      \sb { \seq_item:Nn \l_simone_myq_var_seq { 2 } , #3 }
     }
   }
  ( \lambda\IfValueT{#4}{\sb{#4}} )
 }

\seq_new:N \l_simone_myq_var_seq

\ExplSyntaxOff

\begin{document}

\begin{gather}
\myq{Q} \\
\myq[\bar]{Q} \\
\myq{Q}[i] \\
\myq{Q}(j) \\
\myq{Q}[i](j) \\
\myq[\tilde]{Q}(j) \\
\myq[\tilde]{Q}[i](j) \\
\myq{Q_c} \\
\myq[\bar]{Q_c} \\
\myq{Q_c}[i] \\
\myq{Q_c}(j) \\
\myq{Q_c}[i](j) \\
\myq[\tilde]{Q_c}(j) \\
\myq[\tilde]{Q_c}[i](j)
\end{gather}

\end{document}

enter image description here

egreg
  • 1,121,712