Removing double negative signs in equations

Question

Intro:

I am attempting to write a macro which can remove any extra negative signs in a general equation:

15.2 + -3.1

would be changed to:

15.2 - 3.1

when using the commands:

\newcommand{\A}{15.2}
\newcommand{\B}{-3.1}
\[ A \pluss{\B} \]

My problems start when defining the macro \pluss{} for cases such as:

\newcommand{\A}{5}
\newcommand{\B}{\frac{-3}{12}}
\[ A \pluss{\B} \]

Question:

I thought that in the pluss#1{} macro I would just count the number of negative signs in #1. If it is odd then the result is negative but I am having two problems:

How to count negative signs when the input to the macro is \frac{-3}{12}
How to convert the catcode of the negative signs in \frac{-3}{12} so that they don't print...

Welcome to TeX.SE! Does the fraction have to be stated as \frac{-3}{4} in the \B macro, or would it be OK to state it as -\frac{3}{4}? Could a minus sign be tucked away in the denominator as well? — Mico, Jun 30 '12 at 00:17
Can you create a list of test cases that you would like this to work for. I have an idea, but need to know what kind of expressions you are thinking of and to test that it works for the cases you are thinking of. — Peter Grill, Jun 30 '12 at 01:21
The assumption is also that you don't want two negatives to turn into a positive "automagically", right? — Werner, Jun 30 '12 at 01:25
These are all definitions of \B: -5.7, \alpha, -\alpha, -2.3\beta, -\frac{2}{3}, \frac{-2}{3}, \frac{2}{-3}, \frac{-2}{-3}, -\frac{-2}{-3}. In the end, I do want to output only one + or - symbol. — JAD, Jun 30 '12 at 01:48
@JAD: please just add those to the code above in the form of a MWE. — Peter Grill, Jun 30 '12 at 03:14

score 9 · Answer 1 · answered Jun 30 '12 at 08:26

The argument to \pluss is scanned twice, one for counting the minus signs, the second time ignoring them:

\documentclass{article}

\newcommand{\pluss}{\begingroup\mathcode`-="8000 \plussaux}
\mathchardef\mathminus=\mathcode`-
\newcommand{\plussaux}[1]{%
  \sbox0{\global\minuses=0 \minuscounts$#1$}%
  \ifodd\minuses\mathminus\else+\fi
  \minusignored#1\endgroup}
\newcount\minuses
\def\minuscounts{%
  \begingroup\lccode`~=`- \lowercase{\endgroup
  \def~}{\global\advance\minuses1 }}
\def\minusignored{%
  \begingroup\lccode`~=`- \lowercase{\endgroup
  \def~}{}}

\begin{document}

\parbox{8cm}{
\newcommand\test[1]{\texttt{\detokenize{#1}}\hfill#1}
\linespread{1.5}\selectfont\parfillskip=0pt

\test{$1.2\pluss{5.7}$}

\test{$1.2\pluss{-5.7}$}

\test{$1.2\pluss{\alpha}$}

\test{$1.2\pluss{-\alpha}$}

\test{$1.2\pluss{-2.3\beta}$}

\test{$1.2\pluss{-\frac{2}{3}}$}

\test{$1.2\pluss{\frac{-2}{3}}$}

\test{$1.2\pluss{\frac{2}{-3}}$}

\test{$1.2\pluss{\frac{-2}{-3}}$}

\test{$1.2\pluss{-\frac{-2}{-3}}$}

}

\end{document}

enter image description here

The argument to \pluss should be a monomial expression: something like \pluss{1-2} would produce a weird result.

score 4 · Answer 2 · edited Jun 10 '20 at 12:32

Here is a solution that redefines the \frac macro within the \pluss macro and evaluates the ratio to test if it is negative.

This seems to work for all the cases I could come up with. I attempted to format the output so that it clearly shows the input and output, but there are a few cases (noted with a red asterix) where you need to refer to the code to actually see what it is.

enter image description here

References:

How to test if a number is negative.

Code:

\documentclass{article}
\usepackage{etoolbox}
\usepackage{pgf}
\usepackage{amsmath}
\usepackage{xcolor}
\let\OldFrac\frac
\newtoggle{EncouteredFrac}
\newcommand{\Neumerator}{}%
\newcommand{\Denominator}{}%
\newsavebox\ExpresionBox
\newcommand\pluss[1]{%
    \renewcommand{\frac}[2]{%
        \global\toggletrue{EncouteredFrac}%
        \xdef\Neumerator{##1}%
        \xdef\Denominator{##2}%
    }%
    \global\togglefalse{EncouteredFrac}%
    \sbox\ExpresionBox{\ensuremath{#1}}% set EncountredFraction
    %
    \iftoggle{EncouteredFrac}{%
        \renewcommand{\frac}[2]{(##1 / ##2)}%
        \pgfmathsetmacro{\Var}{#1}%
        \pgfmathtruncatemacro{\AbsNeumerator}{abs(\Neumerator)}%
        \pgfmathtruncatemacro{\AbsDenominator}{abs(\Denominator)}%
        \pgfmathparse{ifthenelse(\Var>=0,"+","-")}%
        \pgfmathresult\OldFrac{\AbsNeumerator}{\AbsDenominator}%
    }{%
        \pgfmathsetmacro{\Var}{#1}%
        \pgfmathsetmacro{\AbsVar}{abs(\Var)}%
        \pgfmathparse{ifthenelse(\Var>=0,"+","-")}%
        \pgfmathresult\AbsVar%
    }%
}%
\newcommand{\A}{15.2}
\newcommand{\PositiveNumber}{3.1}
\newcommand{\NegativeNumber}{-3.1}
\newcommand{\PositiveFracA}{\frac{1}{2}}
\newcommand{\PositiveFracB}{\frac{-1}{-2}}
\newcommand{\NegativeFracA}{\frac{-3}{4}}
\newcommand{\NegativeFracB}{\frac{3}{-4}}
\newcommand{\Note}{\textcolor{red}{$\ast$}}%
\newcommand{\Test}[2]{$#1~\text{\texttt{\textbackslash pluss}} {#2} \to #1 \pluss{#2}$}%
\begin{document}
\section*{These  work}
\par\noindent
\textit{Numbers:}\par
\Test{\A}{3.1}\par
\Test{\A}{-3.1}
\bigskip\par\noindent
\textit{Macros:}\par
\Test{\A}{\PositiveNumber}\par
\Test{\A}{-\PositiveNumber} \Note\quad Subtract a postive number\par
\Test{\A}{\NegativeNumber}  \Note\quad Add a negative number\par
\Test{\A}{-\NegativeNumber} \Note\quad Subtract a negative number\par
\bigskip\par\noindent
\textit{Add a positive frac:}\par
\Test{\A}{\PositiveFracA}\medskip\par
\Test{\A}{\PositiveFracB}
\bigskip\par\noindent
\textit{Add a negative frac:}\par
\Test{\A}{\NegativeFracA}\medskip\par
\Test{\A}{\NegativeFracB}
%----------------------------------
\bigskip\par\noindent
\textit{Subtract a positive frac:}\par
\Test{\A}{-\PositiveFracA}\medskip\par
\Test{\A}{-\PositiveFracB}
\bigskip\par\noindent
\textit{Subtract a negative frac:}\par
\Test{\A}{-\NegativeFracA}\medskip\par
\Test{\A}{-\NegativeFracB}
\end{document}

Removing double negative signs in equations

Intro:

Question:

2 Answers2

References:

Code: