I am using overleaf. I have this block of code written for a 2 column document that needs to be transferred to a single column document.
\begin{figure}
\centering
\begin{minipage}{0.2\textwidth}
\def\xspace{0}
\centering
\hspace{-6mm}
\begin{tikzpicture}
\node[anchor=south west,inner sep=0] (image) at (0,0) {\includegraphics[width=0.8\textwidth,trim={0 0 0 0},clip]{Chapters/ICRA2022/figures/table_items.png}};
\begin{scope}[x={(image.south east)},y={(image.north west)}]
\foreach \i/\j in {{(0.2,0.6)/(0.2,1.1)},{(0.3,0.75)/(0.3,1.1)},{(0.39,1)/(0.39,1.1)},{(0.62,0.7)/(0.62,1.1)}} \draw [-stealth, red, thick] \i -- \j;
\node[scale=0.45] at (0.42,1.15) {$\{O^{(1)},O^{(2)},O^{(3)}, \ ... \ , O^{(7)}\}$};
\node[scale=0.45] at (0.42,1.25) {$O^{(n)}$ for Pick Action};
\end{scope}
\end{tikzpicture}\\
\hspace{-9mm}
\subfloat[]{\begin{tikzpicture}
\node[anchor=south west,inner sep=0] (image) at (0,0) {\includegraphics[width=0.8\textwidth,trim={0 0 0 0},clip]{Chapters/ICRA2022/figures/table_grid.png}};
\begin{scope}[x={(image.south east)},y={(image.north west)}]
\foreach \i/\j in {{(0.015,0.68)/(0.015,1.1)},{(0.15,0.7)/(0.15,1.1)},{(0.285,0.72)/(0.285,1.1)},{(0.42,0.74)/(0.42,1.1)},{(0.557,0.77)/(0.557,1.1)}} \draw [-stealth, red, thick] \i -- \j;
\node[scale=0.45] at (0.33+\xspace,1.15) {$\{O^{(1)},\ O^{(2)},\ O^{(3)},\ O^{(4)}, \ O^{(5)}, \ ... \}$};
\node[scale=0.45] at (0.35+\xspace,1.25) {$O^{(n)}$ for Place Action};
\end{scope}
\end{tikzpicture}}
\end{minipage}\hfill
\begin{minipage}{0.28\textwidth}
\centering
\hspace{-10mm}
\subfloat[]{\includegraphics[width=1.1\textwidth,trim={1cm 0.32cm 0 0.25cm},clip]{Chapters/ICRA2022/figures/architecturenewest.pdf}}
\end{minipage}\caption{(a) $O^{(n)}$ for pick and place actions (b) Our gaze network. The gaze attention block attends to fixations relative to each object while the object attention block attends to objects that are most likely being fixated on.}\label{fig:architecture}
\end{figure}
This is how it originally looks like in the 2 column document.
This is how it looks like in the single column document with too much space in between.
How do I scale the contents within the block of code? Varying the size of the minipage does not work as it will result in the figures looking like this where the text is no longer aligned to the arrows. I do not want to update the positioning manually as I have over 50+ of such figures.
I am looking for a function such as \scale_everything
\begin{figure}
\centering
% % % % % % % % % % % % % % % % % % %
% EXAMPLE USAGE OF SCALE EVERYTHING %
% % % % % % % % % % % % % % % % % % %
\scale_everything{0.5\textwidth}{
\begin{minipage}{0.2\textwidth}
\def\xspace{0}
\centering
\hspace{-6mm}
\begin{tikzpicture}
\node[anchor=south west,inner sep=0] (image) at (0,0) {\includegraphics[width=0.8\textwidth,trim={0 0 0 0},clip]{Chapters/ICRA2022/figures/table_items.png}};
\begin{scope}[x={(image.south east)},y={(image.north west)}]
\foreach \i/\j in {{(0.2,0.6)/(0.2,1.1)},{(0.3,0.75)/(0.3,1.1)},{(0.39,1)/(0.39,1.1)},{(0.62,0.7)/(0.62,1.1)}} \draw [-stealth, red, thick] \i -- \j;
\node[scale=0.45] at (0.42,1.15) {${O^{(1)},O^{(2)},O^{(3)}, \ ... \ , O^{(7)}}$};
\node[scale=0.45] at (0.42,1.25) {$O^{(n)}$ for Pick Action};
\end{scope}
\end{tikzpicture}\
\hspace{-9mm}
\subfloat[]{\begin{tikzpicture}
\node[anchor=south west,inner sep=0] (image) at (0,0) {\includegraphics[width=0.8\textwidth,trim={0 0 0 0},clip]{Chapters/ICRA2022/figures/table_grid.png}};
\begin{scope}[x={(image.south east)},y={(image.north west)}]
\foreach \i/\j in {{(0.015,0.68)/(0.015,1.1)},{(0.15,0.7)/(0.15,1.1)},{(0.285,0.72)/(0.285,1.1)},{(0.42,0.74)/(0.42,1.1)},{(0.557,0.77)/(0.557,1.1)}} \draw [-stealth, red, thick] \i -- \j;
\node[scale=0.45] at (0.33+\xspace,1.15) {${O^{(1)},\ O^{(2)},\ O^{(3)},\ O^{(4)}, \ O^{(5)}, \ ... }$};
\node[scale=0.45] at (0.35+\xspace,1.25) {$O^{(n)}$ for Place Action};
\end{scope}
\end{tikzpicture}}
\end{minipage}}\hfill
\scale_everything{0.4\textwidth}{
\begin{minipage}{0.28\textwidth}
\centering
\hspace{-10mm}
\subfloat[]{\includegraphics[width=1.1\textwidth,trim={1cm 0.32cm 0 0.25cm},clip]{Chapters/ICRA2022/figures/architecturenewest.pdf}}
\end{minipage}}\caption{(a) $O^{(n)}$ for pick and place actions (b) Our gaze network. The gaze attention block attends to fixations relative to each object while the object attention block attends to objects that are most likely being fixated on.}\label{fig:architecture}
\end{figure}
Here is a MWE:
\documentclass{article}
\usepackage[utf8]{inputenc}
\usepackage{float}
\usepackage{subfig}
\usepackage{graphicx}
\usepackage{tikz}
\begin{document}
\begin{figure}
\centering
\subfloat[][a]{
\begin{tikzpicture}
\node[anchor=south west,inner sep=0] (image) at (0,0) {\includegraphics[width=0.3\textwidth,trim={0 0 0 0},clip]{example-image-a}};
\begin{scope}[x={(image.south east)},y={(image.north west)}]
\foreach \i/\j in {{(0.2,0.6)/(0.2,1.1)},{(0.3,0.75)/(0.3,1.1)},{(0.39,1)/(0.39,1.1)},{(0.62,0.7)/(0.62,1.1)}} \draw [-stealth, red, thick] \i -- \j;
\node[scale=0.45] at (0.42,1.15) {${O^{(1)},O^{(2)},O^{(3)}, \ ... \ , O^{(7)}}$};
\node[scale=0.45] at (0.42,1.25) {$O^{(n)}$ for Pick Action};
\end{scope}
\end{tikzpicture}}
\subfloat[][a]{
\begin{tikzpicture}
\node[anchor=south west,inner sep=0] (image) at (0,0) {\includegraphics[width=0.3\textwidth,trim={0 0 0 0},clip]{example-image-a}};
\begin{scope}[x={(image.south east)},y={(image.north west)}]
\foreach \i/\j in {{(0.2,0.6)/(0.2,1.1)},{(0.3,0.75)/(0.3,1.1)},{(0.39,1)/(0.39,1.1)},{(0.62,0.7)/(0.62,1.1)}} \draw [-stealth, red, thick] \i -- \j;
\node[scale=0.45] at (0.42,1.15) {${O^{(1)},O^{(2)},O^{(3)}, \ ... \ , O^{(7)}}$};
\node[scale=0.45] at (0.42,1.25) {$O^{(n)}$ for Pick Action};
\end{scope}
\end{tikzpicture}}
\label{steady_state}
\end{figure}
\end{document}




\scalebox{2}{...}, but this is really bad design. You do not want to hardcode coordinates to point at text and you do also not want to scale text like this. Text font sizes are not just scales and text should be consistent throughout the document. – hpekristiansen Oct 15 '22 at 20:32