# label a part of commutative diagram

\documentclass[12pt,reqno,a4paper]{amsart}
\usepackage{extsizes}
\usepackage{blindtext}
\textheight 9.3in \textwidth 6.5in
\calclayout
\usepackage{amsmath,amsthm,amsfonts,amssymb}
\usepackage{hyperref}
\usepackage{mathrsfs}
\usepackage[all]{xy}
\usepackage[normalem]{ulem}
\usepackage{tikz-cd}
\usepackage[utf8]{inputenc}
\usepackage[english]{babel}
\begin{document}
$\begin{tikzcd} \underline{V} \arrow[rrrrdd, bend left] \arrow[rrdddd, bend right] \arrow[rrdd, dotted] & && & \\ & & & & \\ & & A \arrow[dd] \arrow[rr] & & B \arrow[dd] \\ & & & & \\ & & C \arrow[rr] & & D \end{tikzcd}$
\end{document}


above code is giving me following diagram

I used "phantom" technique to number part of commutative diagram. It did not work. I want to write

• (1) in the diagram enclosed by V,A,B
• (2) in the diagram enclosed by V,A,C
• (3) in the diagram enclosed by A,B,C,D

Can some one help me to see how to do this?

Is this what you mean?

$\begin{tikzcd} \underline{V} \arrow[rrrrdd, bend left] \arrow[rrdddd, bend right] \arrow[rrdd, dotted] & && & \\ & & (1) & & \\ & (2) & A \arrow[dd] \arrow[rr] & & B \arrow[dd] \\ & & & (3) & \\ & & C \arrow[rr] & & D \end{tikzcd}$

• absolutely.. this seem to be much easier than phantom thing,.. Mar 18 '19 at 12:40

Just for fun: tikz-cd is based on TikZ so

\documentclass[tikz]{standalone}
\begin{document}
\begin{tikzpicture}[y=0.8cm]
\node (a) at (0,0) {$A$};
\node (b) at (2,0) {$B$};
\node (c) at (0,-2) {$C$};
\node (d) at (2,-2) {$D$};
\node (v) at (-2,2) {$\underline{V}$};
\draw[->] (a)--(b);
\draw[->] (b)--(d);
\draw[->] (a)--(c);
\draw[->] (c)--(d);
\draw[dotted,->] (v)--(a);
\draw[->] (v) to[bend right] (c);
\draw[->] (v) to[bend left] (b);
\node at (1,-1) {(3)};
\node at (0,1) {(1)};
\node at (-1,0) {(2)};
\end{tikzpicture}
\end{document}


But I prefer this way (though the way I code it is not efficient)

\documentclass[tikz]{standalone}
\begin{document}
\begin{tikzpicture}[y=0.8cm,>=stealth]
\node (V) at (-2,2) {\phantom{$\underline{V}$}};
\node (B) at (2,0) {\phantom{$B$}};
\node (C) at (0,-2) {\phantom{$C$}};
\fill[yellow!50] (0,0) rectangle (2,-2);
\fill[green!20] (V) to[bend left] (B)--(2,0)--(0,0)--(V);
\fill[blue!20] (V) to[bend right] (C)--(0,-2)--(0,0)--(V);
\begin{scope}[every node/.style={fill=white,circle,draw}]
\node (a) at (0,0) {$A$};
\node (b) at (2,0) {$B$};
\node (c) at (0,-2) {$C$};
\node (d) at (2,-2) {$D$};
\node (v) at (-2,2) {$\underline{V}$};
\end{scope}
\draw[->] (a)--(b);
\draw[->] (b)--(d);
\draw[->] (a)--(c);
\draw[->] (c)--(d);
\draw[dashed,->] (v)--(a);
\draw[->] (v) to[bend right] (c);
\draw[->] (v) to[bend left] (b);
\node at (1,-1) {(3)};
\node at (0,1) {(1)};
\node at (-1,0) {(2)};
\end{tikzpicture}
\end{document}


• :) :) thanks... Mar 18 '19 at 12:41
• @PraphullaKoushik You are welcome. Now you have got three answers, choose the best of them and tick the checkmark on the left of the answer (aka accept that answer). Doing so will mark this question as resolved.
– user156344
Mar 19 '19 at 5:32

With phantom arrows:

\documentclass[a4paper]{article}
\usepackage{tikz-cd}

\begin{document}

$\begin{tikzcd}[nodes in empty cells] \underline{V} \arrow[rrrrdd, bend left] \arrow[rrdddd, bend right] \arrow[rrdd, dotted] \arrow[rrrrdd,phantom,"(1)" description] \arrow[rrdddd,phantom,"(2)" description] \\ &&&&\\ && A \arrow[dd] \arrow[rr] & \arrow[dd,phantom,"(3)" description] & B \arrow[dd] \\ &&&&\\ && C \arrow[rr] && D \end{tikzcd}$

\end{document}