# Checking if a number is a multiple of 4

Using etoolbox, ifthen, or another form of conditional checking, how can I check if a value is a multiple of 4? E.g. if the number is 4, 8, 12, 16, etc., then it returns "true".

-

You can use pgfmath for this:

\documentclass{article}
\usepackage{pgf}

\begin{document}
\newcommand\ifismultiple[4]{%
\pgfmathparse{mod(#1,#2)==0}
\ifnum \pgfmathresult=1
#3%
\else
#4%
\fi
}
\ifismultiple{15}{4}{true}{false}
\end{document}

-
If you swap the first two arguments, then you can define a new command \newcommand\ifismultipleoffour{\ifismultiple{4}}, etc. – Aditya Dec 19 '11 at 5:01

\ifodd is enough to do this

\documentclass[]{scrartcl}

\newcount\myint

\def\IsFourMultiple#1{%
\myint#1\relax
\ifodd\myint
false
\else
\divide\myint by2\relax
\ifodd\myint
false
\else
true
\fi
\fi}%

\begin{document}

\IsFourMultiple{4}
\IsFourMultiple{32}
\IsFourMultiple{15}

\end{document}

-
Ah, that's very clever! – Jake Dec 19 '11 at 6:05
@Jake But my first idea was to use pgfmath' like you, because I always have \usepackage{tikz} in my preamble. Your solution is more general. – Alain Matthes Dec 19 '11 at 6:44
+1 for cleverness – Marco Daniel Dec 19 '11 at 7:57
One drawback (shared by many of the solutions here) is that it is not expandable. – Bruno Le Floch Dec 19 '11 at 12:39
@BrunoLeFloch Yes but it is sometimes a desirable and useful feature but it's not always desirable. Now I would like to know what is the cost in terms of efficiency. – Alain Matthes Dec 19 '11 at 13:05

With eTeX's \numexpr, you can do that easily.

\makeatletter
\newcommand{\doifmultipleof}[2]{%
\ifnum\numexpr((#2)/(#1))*(#1)-(#2)=0
\expandafter\@firstoftwo
\else
\expandafter\@secondoftwo
\fi}
\makeatother

\documentclass{article}
\begin{document}
\doifmultipleof{4} {6}{true}{false} % => false
\doifmultipleof{5}{10}{true}{false} % => true
\end{document}

-
One should note that e-TeX's division rounds, so this method is not suitable to compute the remainder. This doesn't invalidate this solution, of course, as the difference between a multiple of 4 and a non-multiple of 4 can never be 0. – egreg Dec 19 '11 at 10:18
@egreg. Yes, and the LaTeX3 function \int_mod:nn uses some very clever code (by Heiko I believe) to work around that, while avoiding overflow in all cases. – Bruno Le Floch Dec 19 '11 at 12:38

If you can use luatex, then the comparison can be easily done at the lua end. Here is a ConTeXt solution (using named macro arguments):

% macros=mkvi

\def\doifmultipleof#{mod}#{num}%
{\ctxcommand{doifelse(math.mod(#{num}, #{mod}) == 0)}}

\starttext
\dorecurse{10}
{\doifmultipleof{4}{\recurselevel}{yes}{no}}
\stoptext

-
+1 because this is the only expandable answer so far (it is expandable, right?). But I have to say, I find the named macro argument syntax quite outlandish. :) – Bruno Le Floch Dec 19 '11 at 5:02
@BrunoLeFloch: Yes, the code is expandable. As for the named arguments, one can also use #mod and #num; in fact, those are the recommended alternatives, but I prefer the #{...} variants. – Aditya Dec 19 '11 at 6:09
just curious: with the #mod variant, how is the end of the argument name detects? Can you have delimited arguments that way? – Bruno Le Floch Dec 19 '11 at 6:31
\documentclass{minimal}
\def\IsFourMultiple#1{%
\ifdim\dimexpr#1pt/4=\numexpr#1/4pt true\else false\fi}
\begin{document}

\IsFourMultiple{4}
\IsFourMultiple{32}
\IsFourMultiple{15}

\end{document}

-
\IsFourMultiple{2147483647} – Andrey Vihrov Dec 19 '11 at 8:21
every solution has a limit! not really surprising ... ;-) – Herbert Dec 19 '11 at 9:15
Interesting approach. If eTeX is available, you'd better do \ifdim\dimexpr#1pt/4=\numexpr#1/4pt : that makes your macro expandable, and probably a tiny bit faster. – Bruno Le Floch Dec 19 '11 at 13:37
true, thanks ... – Herbert Dec 19 '11 at 16:09

The fp package can be used to calculate stuff and perform modular arithmetic. Here's a minimal example:

\documentclass{article}
\usepackage{etoolbox}% http://ctan.org/pkg/etoolbox
\usepackage[nomessages]{fp}% http://ctan.org/pkg/fp
\newcommand{\multoffour}[1]{%
\FPeval{\result}{trunc(#1-(4*trunc(#1/4,0)),0)}% \result contains #1 mod 4
\ifnumequal{\result}{0}{true}{false}% true = multiple of four; false = not a multiple of four
}
\begin{document}
1 is a multiple of~4: \multoffour{1} \par
2 is a multiple of~4: \multoffour{2} \par
3 is a multiple of~4: \multoffour{3} \par
4 is a multiple of~4: \multoffour{4} \par
5 is a multiple of~4: \multoffour{5} \par
6 is a multiple of~4: \multoffour{6} \par
7 is a multiple of~4: \multoffour{7} \par
8 is a multiple of~4: \multoffour{8} \par
9 is a multiple of~4: \multoffour{9} \par
\end{document}


etoolbox was used for numeric comparisons.

This works just as well with counters. For example, using

\newcounter{mycntr} \setcounter{mycntr}{10}
\multoffour{\themycntr}


(provided that \themycntr is defined using \arabic{mycntr} or not modified).

-
Isn't fp package (or, for that matter, any package) an overkill for this task? And is this fp solution expandable? fp solutions are hardly expandable. For non-context and non-lua users, I think Bruno's solution is more than enough, IMHO. And it's superfast. I often get scared by fp's trace log :) – Ahmed Musa Dec 19 '11 at 16:46
@AhmedMusa: Sure, there are at least three other answers that produce a package-less solution. Honestly, speed for me is not an issue, since the time difference between fp's output and some other package-less output is most likely less than 0.1 seconds. I rely on code readability, which fp provides through its function shorthand/names. If I (or anyone for that matter) is relying on performing major calculations within TeX and want speed, they are perhaps better off using a different environment altogether. ;)` – Werner Dec 19 '11 at 16:54