# DAVIDSTUTZ

Check out the latest superpixel benchmark — Superpixel Benchmark (2016) — and let me know your opinion! @david_stutz
03rdNOVEMBER2016

## Simple Line Plots in LaTeX using PGFPlots

A simple example of line plots in LaTeX using PGFPlots. The example is part of a collection of LaTeX snippets on GitHub.

line-plots.tex
\documentclass[11pt]{article}

\usepackage{amsmath}
\usepackage{amssymb}
\usepackage{subfigure}

\usepackage[font=footnotesize]{caption}
\usepackage{pgfplots}
\usepackage{tikz}

% Scriptsize axis style.
\pgfplotsset{every axis/.append style={tick label style={/pgf/number format/fixed},font=\scriptsize,ylabel near ticks,xlabel near ticks,grid=major}}

\begin{document}
\begin{figure}[t]
\centering
\subfigure[USPS]{
\begin{tikzpicture}
\begin{axis}[%
xlabel=\% of Dataset,
ylabel=Accuracy in \%,
height=5.5cm,
width=6cm,
ymax=96,
ymin=28,
xmin=0,
xmax=100,
%xmode=log,
%log ticks with fixed point,
]

% T = 100
(0.5,38.76)
(1,65.32)
(2.5,75.19)
(5,79.32)
(10,85.10)
(25,89.04)
(50,90.13)
(100,91.43)
};

(0.5,29.65)
(1,40.81)
(2.5,64.23)
(5,65.42)
(10,76.03)
(25,83.41)
(50,87.44)
(100,89.49)
};

% T = 10
(0.5,47.93)
(1,55.71)
(2.5,71.70)
(5,78.48)
(10,81.96)
(25,86.75)
(50,87.59)
(100,90.28)
};

(0.5,36.07)
(1,23.47)
(2.5,44.69)
(5,42.85)
(10,65.37)
(25,76.83)
(50,80.97)
(100,83.31)
};

% T = 1
(0.5,24.36)
(1,35.33)
(2.5,40.36)
(5,59.04)
(10,62.48)
(25,68.81)
(50,71.30)
(100,78.08)
};

(0.5,17.89)
(1,23.47)
(2.5,27.01)
(5,38.12)
(10,38.47)
(25,45.47)
(50,54.71)
(100,57.60)
};
\end{axis}
\end{tikzpicture}
}
\subfigure[MNIST]{
\begin{tikzpicture}
\begin{axis}[%
xlabel=\% of Dataset,
ylabel=Accuracy in \%,
height=5.5cm,
width=6cm,
ymax=96,
ymin=28,
xmin=0,
xmax=100,
%xmode=log,
%log ticks with fixed point,
]

% T = 100
(0.5,79.58)
(1,85.22)
(2.5,89.69)
(5,91.09)
(10,92.96)
(25,94.10)
(50,95.02)
(100,95.64)
};
\label{plot:mnist-100-offline}

(0.5,64.80)
(1,73.88)
(2.5,80.98)
(5,84.61)
(10,89.37)
(25,91.15)
(50,92.60)
(100,93.61)
};
\label{plot:mnist-100-online}

% t = 10
(0.5,67.65)
(1,76.72)
(2.5,83.91)
(5,87.57)
(10,89.95)
(25,91.40)
(50,92.85)
(100,89.21)
};
\label{plot:mnist-10-offline}

(0.5,54.90)
(1,60.64)
(2.5,69.93)
(5,74.11)
(10,80.60)
(25,84.79)
(50,86.97)
(100,89.21)
};
\label{plot:mnist-10-online}

% T = 1
(0.5,39.98)
(1,49.64)
(2.5,57.94)
(5,61.12)
(10,68.32)
(25,70.45)
(50,73.37)
(100,78.96)
};
\label{plot:mnist-1-offline}

(0.5,25.15)
(1,37.37)
(2.5,34.70)
(5,43.81)
(10,44.85)
(25,51.46)
(50,56.54)
(100,61.47)
};
\label{plot:mnist-1-online}
\end{axis}
\end{tikzpicture}
}
\caption{Comparison of on-line and off-line Random Forests on the USPS \cite{Hull} and MNIST \cite{LeCunBottouBengioHaffner} datasets. Note that the MNIST dataset provides $60,000$ training examples and $10,000$ test examples with dimension $28 \times 28$., while the USPS dataset only provides $7,291$ training examples and $2,007$ test examples with dimension $16 \times 16$. Experiments where conducted for different forest sizes $T$: \ref{plot:mnist-1-offline}, \ref{plot:mnist-1-online} $T = 1$; \ref{plot:mnist-10-offline}, \ref{plot:mnist-10-online} $T = 10$; and \ref{plot:mnist-100-offline}, \ref{plot:mnist-100-online} $T = 100$}
\label{fig:online-random-forests-experiments}
\end{figure}

\begin{thebibliography}{1}
\bibitem{Hull}
J. J. Hull,
\emph{A Database for Handwritten Text Recognition Research},
Transactions on Pattern Analysis and Machine Intelligence,
1994.

\bibitem{LeCunBottouBengioHaffner}
Y. LeCun,
L. Bottou,
Y. Bengio,
P. Haffner,
\emph{Gradient-Based Learning Applied to Document Recognition},
Proceedings of the IEEE,
1998.
\end{thebibliography}
\end{document}

line-plots.pdf
line-plots

What is your opinion on the code snippet? Is it working? Let me know your thoughts in the comments below or using the following platforms: