\newcommand in LaTeX to return a TikZ node

Question

In order to avoid repetitive lines of code, we have\tikzset.

On similar lines , is there a possibility to create nodes with \newcommand.

For example,

Instead of \node[draw=black] (a) at (0,0) {} and \node[draw=black] (b) at (0,5) {} with two node statements , We can use single statement,

\newcommand[2]{\createnode}{\node[draw=black] #1 at #2 {}}

Will this approach work and will the nodes created via this approach be accessible at a later stage?

Also with this approach, how will scope work if we create the same node in two scopes.

Will this work?

Sorry, don't mean to be rude, but why don't you try this and ask the follow-up question in case it turns out not to work? Feels like the kind of problem you could try in less time than it takes for us to read that question. — sheß, Apr 01 '19 at 17:25
TikZ has all the means to avoid using \newcommands. \documentclass[border=3.14mm,tikz]{standalone} \begin{document} \begin{tikzpicture}[soni/.style n args={2}{insert path={node[draw=black] #1 at #2 {}}}] \path[soni={(a)}{(0,0)}]; \end{tikzpicture} \end{document} — , Apr 01 '19 at 18:07
Using \newcommand as you suggest works as expected. I have used that often and have not had any major issues with it. Used within a scope also works exactly as expected. The only minor issue I am aware of (which is not an issue with tikz) is with the use of trailing ; following the use of a macro is the nullfont warnings in the log file. This is explained nicely by Martin Scharrer's comment at [tikzpicture options causing “There is no in font nullfont!”?. But, if you are aware of this upfront you can easily avoid it. — Peter Grill, Apr 01 '19 at 18:31

score 2 · Answer 1 · answered Apr 04 '19 at 07:28

First of all, your syntax of \newcommand is wrong.

\newcommand[2]{\createnode}{\node[draw=black] #1 at #2 {}}

won't work. These ones will work

\newcommand{\createnode}[2]{\node[draw=black] #1 at #2 {}}
\newcommand\createnode[2]{\node[draw=black] #1 at #2 {}} % I prefer this way

Regarding the first question,

Will this approach work and will the nodes created via this approach be accessible at a later stage?

It works (see the example below), but take care of the braces, brackets and semicolons.

\documentclass[tikz]{standalone}
\newcommand\createnode[2]{\node[draw=black] (#1) at (#2) {}}
\begin{document}
\begin{tikzpicture}
\createnode{x}{0,0};
\createnode{y}{3,4};
\draw[-stealth] (x) to[bend left] (y);
\end{tikzpicture}
\end{document}

It can be used with normal \nodes:

\documentclass[tikz]{standalone}
\newcommand\createnode[2]{\node[draw=black] (#1) at (#2) {}}
\begin{document}
\begin{tikzpicture}
\createnode{x}{0,0};
\node[circle,draw] (y) at (3,4) {$y$};
\draw[-stealth] (x) to[bend left] (y);
\end{tikzpicture}
\end{document}

It can even be used in remember picture

\documentclass[varwidth]{standalone}
\usepackage{tikz}
\newcommand\createnode[2]{\node[draw=black] (#1) at (#2) {}}
\begin{document}
This is a node: \tikz[remember picture]{\createnode{x}{0,2};}

And this is another node: \tikz[remember picture]{\createnode{y}{1,3};}

Let's try adding an arrow between them. It works!
\begin{tikzpicture}[overlay,remember picture]
\draw[-stealth] (x) to[bend left] (y);
\end{tikzpicture}
\end{document}

Regarding the second question,

Also with this approach, how will scope work if we create the same node in two scopes.

Consider the following code: it is not using \createnode:

\documentclass[tikz]{standalone}
%\newcommand\createnode[2]{\node[draw=black] (#1) at (#2) {}}
\begin{document}
\begin{tikzpicture}
\begin{scope}
\node[draw] (x) at (2,3) {};
\node[draw] (z) at (0,3) {};
\draw[-stealth] (z) -- (x);
\end{scope}
\begin{scope}
\node[draw] (x) at (0,0) {};
\node[draw] (y) at (2,1) {};
\end{scope}
\draw[-stealth] (z) -- (x);
\end{tikzpicture}
\end{document}

Node (x) is redefined in the later scope, so the two arrows are different.

Let try it with \createnode:

\documentclass[tikz]{standalone}
\newcommand\createnode[2]{\node[draw=black] (#1) at (#2) {}}
\begin{document}
\begin{tikzpicture}
\begin{scope}
\node[draw] (x) at (2,3) {};
\node[draw] (z) at (0,3) {};
\draw[-stealth] (z) -- (x);
\end{scope}
\begin{scope}
%\node[draw] (x) at (0,0) {};
\createnode{x}{0,0};
\node[draw] (y) at (2,1) {};
\end{scope}
\draw[-stealth] (z) -- (x);
\end{tikzpicture}
\end{document}

\documentclass[tikz]{standalone}
\newcommand\createnode[2]{\node[draw=black] (#1) at (#2) {}}
\begin{document}
\begin{tikzpicture}
\begin{scope}
\node[draw] (x) at (2,3) {};
\createnode{x}{2,3};
\node[draw] (z) at (0,3) {};
\draw[-stealth] (z) -- (x);
\end{scope}
\begin{scope}
\node[draw] (x) at (0,0) {};
%\createnode{x}{0,0};
\node[draw] (y) at (2,1) {};
\end{scope}
\draw[-stealth] (z) -- (x);
\end{tikzpicture}
\end{document}

Actually nothing changes.

My conclusion: If you use it carefully, it will definitely work as expected.

My comments: I think defining a new command for this is quite good. Sometimes I use this way too. However, please only use it when your command is really complicated and it is repeated several times, because you have to use it very carefully – the lack of a single semicolon can cause you many troubles. Don't overuse this way for simple commands like \node, otherwise it may be very hard to debug your code in some cases.

\newcommand in LaTeX to return a TikZ node

1 Answers1