Using TikZ/PGF to Create LaTeX Graphics

Info

This guide uses tikzjax to render TikZ/PGF code in the browser. If you cannot see the graphics, try refreshing the page. If that does not work you may need to enable JavaScript or view the page in a browser that supports JavaScript.

TikZ and PGF can produce highly complex code and graphics, but in doing so, they require more time and effort to learn and use effectively. This guide will provide a brief overview of how to use TikZ/PGF and PGFPlots to create graphics in LaTeX documents. However, I would strongly recommend reading the following resources for a fantastic guide on how to use them that will be far more comprehensive than this guide:

Unlocking LaTeX Graphics - Dr Tamara G. Kolda
TikZ & PGF Manual
PGFPlots Online Manual - Unofficial HTML manual but very helpful
PGFPlots Manual

Like in Kolda's guide this page will focus on the most used features of TikZ and PGF and most relevant elements to producing graphics in a report, it will not cover every aspect of TikZ and PGF.

If you find you don't want to code all the graphics yourself there are still some helpful tools that can do most of the work for you such as the PGFPlotsEdt tool found here PGFPlotsEdt (Github).

There can be some confusion around TikZ and PGF and how they relate to each other. TikZ is a frontend for PGF, which is a lower-level language for producing graphics in LaTeX. TikZ is a more user-friendly interface for PGF, and it is what most people use when creating graphics in LaTeX. However, PGF is more powerful and can be used to create more complex graphics than TikZ. For the purposes of this guide, we will be using TikZ, but it is worth noting that PGF is also available if you need more power.

There is also PGFPlots which is a package built on top of TikZ and PGF that is specifically designed for creating plots and charts. It features plot types that are not natively available in TikZ/PGF, such as box plots. It is maintained by the same team that maintains TikZ and PGF, so it is well-integrated with the rest of the LaTeX ecosystem.

Additionally, while these use to be separate packages, they are now bundled together in the same package, so you only need to include \usepackage{tikz} in your preamble. But libraries can be loaded with either \usetikzlibrary{<library>} or \usepgflibrary{<library>}.

If you want to use PGFPlots, you will need to include \usepackage{pgfplots} in your preamble.

The following will show you how to create graphics using the datavisulization tool and PGFPlots.

Getting Data in to TikZ

There are a few ways to get data in to TikZ, the most common are:

Hardcoding the data in the TikZ code
Using a CSV file
Using a data file

The most common method is to hardcode the data in the TikZ code, but this can be cumbersome for large datasets. Using a CSV file is a good alternative, as it allows you to keep the data separate from the code. This can be done using:

To read from a file:

\pgfdata[format=table, read from file=<filename.csv>]

\datavisualization [
    scientific axes=clean,
    x axis={length=3cm, ticks=few},
    all axes={grid},
    visualize as smooth line
    ] 
    data [format=table, read from file=file1.csv];

Alternatively you can use pgfplotstable and pgfplots packages to read in data from a file, you will need to specify the column separator:

\usepackage{pgfplots}
\usepackage{pgfplotstable}

\begin{document}

\pgfplotstableread[col sep=comma]{<filename.csv>}\data

\begin{tikzpicture}
\begin{axis}[grid=both]
\addplot[smooth] table {\data};
\end{axis}
\end{tikzpicture}

\end{document}

As you will notice this is a lot more verbose than the datavisualization method. Datavisualization is designed to be simpler and is more user-friendly for most basic plots. However, pgfplots is a very powerful package and is widely used and has a wealth of documentation due to its age while data visualization is still relatively new in the scheme of LaTeX packages.

Note

PGFPlots has not been updated since 2021. While PGF (which contains Datavisualization) is still actively updated. This is not to say that PGFPlots is not a good package, it is still widely use and can plot things that Datavisualization cannot. But it is worth noting that Datavisualization is the newer package and is more user-friendly when a simpler plot is needed.

Note

The following sections will include examples of both pgfplots and datavisualization to show the differences between the two packages, where applicable. Some plots may not be available in datavisualization and will be shown using pgfplots only.

Scatter Plots

Datavisualization

\usetikzlibrary{datavisualization}
\begin{document}
\begin{tikzpicture}
\datavisualization [
    scientific axes={clean, width=.55\textwidth, height=5.5cm},
    x axis={ticks=few},
    y axis={max value=250, min value=-20},
    all axes={grid},
    visualize as scatter/.list={explt},
    text=gray,
    visualizer color=red!90,
    explt={style={red,line width=2pt}}
    ] 
    data [set=explt] {
        x, y
        0, 0
        1, 3
        2, 9
        3, 17
        4, 90
        5, 21
        6, 2
        7, 229
        8, 3
        9, 81
        10, 2
    };
\end{tikzpicture}
\end{document}

PGFPlots

\usepackage{pgfplots}
\begin{document}
\begin{tikzpicture}
\begin{axis}[grid=both, axis line style=gray, text=gray, xmin=0, xmax=10, xtick distance=2.5]
\addplot [only marks, color=red!90, line width=2pt] coordinates {
        (0, 0)
        (1, 3)
        (2, 9)
        (3, 17)
        (4, 90)
        (5, 21)
        (6, 2)
        (7, 229)
        (8, 3)
        (9, 81)
        (10, 2)
    };
\end{axis}
\end{tikzpicture}
\end{document}

Line Plots

Datavisualization

\usetikzlibrary{datavisualization}
\begin{document}
\begin{tikzpicture}
\datavisualization [
    scientific axes={clean, width=.55\textwidth, height=5.5cm},
    x axis={ticks=few},
    y axis={max value=250, min value=-20},
    all axes={grid},
    visualize as smooth line/.list={explt},
    text=gray,
    visualizer color=red!90,
    explt={style={red,line width=2pt}}
    ] 
    data [set=explt] {
        x, y
        0, 0
        1, 3
        2, 9
        3, 17
        4, 90
        5, 21
        6, 2
        7, 229
        8, 3
        9, 81
        10, 2
    };
\end{tikzpicture}
\end{document}

PGFPlots

\usepackage{pgfplots}
\begin{document}
\begin{tikzpicture}
\begin{axis}[grid=both, axis line style=gray, text=gray, xmin=0, xmax=10, xtick distance=2.5]
\addplot [smooth, color=red!90, line width=2pt] coordinates {
        (0, 0)
        (1, 3)
        (2, 9)
        (3, 17)
        (4, 90)
        (5, 21)
        (6, 2)
        (7, 229)
        (8, 3)
        (9, 81)
        (10, 2)
    };
\end{axis}
\end{tikzpicture}
\end{document}

Bar Plots

Note

Shown using pgfplots only.

\usepackage{pgfplots}
\begin{document}
\begin{tikzpicture}
\begin{axis}[grid=both, axis line style=gray, text=gray, xmin=0, xmax=10, xtick distance=2.5]
\addplot [ybar, color=red!90, fill=red!90, fill opacity=0.2, line width=2pt] coordinates {
        (0, 0)
        (1, 3)
        (2, 9)
        (3, 17)
        (4, 90)
        (5, 21)
        (6, 2)
        (7, 229)
        (8, 3)
        (9, 81)
        (10, 2)
    };
\end{axis}
\end{tikzpicture}
\end{document}

Stacked Plots

Note

Shown using pgfplots only.

\usepackage{pgfplots}
\pgfsetblendmode{screen}
\begin{document}
\begin{tikzpicture}
\begin{axis}[
    stack plots=y,
    text=gray, 
    area style,
    enlarge x limits=false,
    smooth,
    fill opacity=0.7,
    grid=both, 
    axis line style=gray,
    xmin=0, 
    xmax=10,    
    xtick distance=2.5
]
    \addplot coordinates
        {(0,1) (1,1) (2,2) (3,2) (4, 20) (5, 21) (6, 2) (7, 19) (8, 20) (9, 1) (10, 2)}
            \closedcycle;
    \addplot coordinates
        {(0,1) (1,3) (2,7) (3,2) (4, 3) (5, 25) (6, 2) (7, 19) (8, 4) (9, 10) (10, 8)}
            \closedcycle;
    \addplot coordinates
        {(0,1) (1,4) (2,2) (3,2) (4, 13) (5, 11) (6, 20) (7, 12) (8, 15) (9, 23) (10, 12)}
            \closedcycle;
    \end{axis}
\end{tikzpicture} 
\end{document}

Box Plots

Note

Shown using pgfplots only.

\usepackage{pgfplots}
\usepgfplotslibrary{statistics}

\begin{document}
    \begin{tikzpicture}
        \begin{axis}[grid=both, axis line style=gray, text=gray, xtick distance=2.5]
            \addplot+ [boxplot, color=red!90, fill=red!90, fill opacity=0.2, line width=2pt] table [row sep=\\,y index=0] {
                data\\
                0\\
                3\\
                9\\
                17\\
                21\\
                2\\
                3\\
                2\\
            };
        \end{axis}
    \end{tikzpicture}
\end{document}

If you would like to use violin plots instead you will need the tikzviolinplots package.

Histograms

Note

Shown using pgfplots only.

\usepackage{pgfplots}
\usepgfplotslibrary{statistics}

\begin{document}
    \begin{tikzpicture}
        \begin{axis}[ybar interval, grid=both, axis line style=gray, text=gray, xtick distance=2.5]
            \addplot+ [hist={bins=5}, color=red!90, fill=red!90, fill opacity=0.2, line width=2pt] table [row sep=\\,y index=0] {
                data\\
                0\\
                3\\
                9\\
                17\\
                21\\
                2\\
                3\\
                2\\
            };
        \end{axis}
    \end{tikzpicture}
\end{document}

Heat Maps

Note

Shown using pgfplots only.

\usepackage{pgfplots}
\begin{document}
\begin{tikzpicture}
\begin{axis}[enlargelimits=false,colorbar, axis line style=gray, text=gray, colormap/viridis, colorbar style={text=gray}]
    \addplot [matrix plot,
        point meta=explicit,
    ] coordinates {
        (0,0) [0] (1,0) [1] (2,0) [2] (3,0) [1] (4,0) [2] (5,0) [1]

        (0,1) [3] (1,1) [43] (2,1) [51] (3,1) [12] (4,1) [2] (5,1) [13]

        (0,2) [36] (1,2) [7] (2,2) [28] (3,2) [31] (4,2) [42] (5,2) [1]

        (0,3) [0] (1,3) [41] (2,3) [2] (3,3) [1] (4,3) [2] (5,3) [9]

        (0,4) [3] (1,4) [4] (2,4) [5] (3,4) [1] (4,4) [32] (5,4) [1]

        (0,5) [6] (1,5) [72] (2,5) [68] (3,5) [1] (4,5) [2] (5,5) [1]
    };
\end{axis}
\end{tikzpicture} 
\end{document}

Contour Plots

Note

Shown using pgfplots only. Requires LuaLaTeX.

\usepackage{pgfplots}
\pgfplotsset{height=5.5cm,compat=1.18}
\begin{document}
    \begin{tikzpicture}
        \begin{axis}[view={0}{90}, axis line style=gray, text=gray]
            \addplot3 [
            contour lua={number=14, labels=false}
            ] {x*y};
        \end{axis}
    \end{tikzpicture} 
\end{document}

3D Plots

Note

Shown using pgfplots only. This can be very resource intensive and may not work on all systems.

\usepackage{pgfplots}
\begin{document}
\begin{tikzpicture}
    \begin{axis}[
        3d box=complete,
        grid=major,
        samples=5, domain=-5:5,
        xtick=data, ytick=data,
        axis line style=gray, 
        text=gray, 
        colormap/viridis
    ]
        \addplot3 [surf] {x*y};
    \end{axis}
\end{tikzpicture}%
\end{document}

PGF Math Capabilities

PGF has extensive math capabilities, and you can use it to create complex mathematical expressions. This is particularly useful if you need to perform operations on data or create complex mathematical diagrams.

For example polynomial approximation can be done with the following code:

\pgfmathveclen{x}{y} \pgfmathresult

This is not a replacement for a full programming language, but it can be useful for simple calculations. More information on the math capabilities of PGF can be found in the TikZ & PGF Manual under section 94.

Note

For more comprehensive mathematical operations and manipulations please refer to Extending LaTeX.

Importing Graphics from MATLAB

Note

I do not use MATLAB and have not tested this method. This is based on the documentation available.

It is possible to import graphics from MATLAB in to LaTeX using matlab2tikz script. This script will convert your MATLAB graphics in to TikZ code that you can then import in to your LaTeX document. This is particularly useful if you are more comfortable with MATLAB and want to create complex graphics in MATLAB and then import them in to LaTeX.

To use matlab2tikz, you will need to install it in MATLAB. You can do this by downloading the script from the Mathworks File exchange matlab2tikz and then adding it to your MATLAB path. You can then use the matlab2tikz function to convert your MATLAB graphics in to TikZ code.

For example, you can create a simple plot in MATLAB:

x = linspace(0,2*pi,100);
y = sin(x);
plot(x,y)

And then convert it to TikZ code using matlab2tikz:

matlab2tikz('test.tex')

TikZ without the TikZ

For those that do not want to learn TikZ, there are a few tools that can be used to create TikZ code without actually writing it. These tools are:

Inkscape
Ipe

This guide will not delve in to the specifics of these tools, but they are worth mentioning for those that do not want to learn TikZ. And good guidance is available in their respective documentation.

Inkscape

Inkscape is a powerful vector graphics editor that can be used to create TikZ code. This is particularly useful for creating complex graphics that would be difficult to create just writing the code by hand. To allow for TikZ code to be exported from Inkscape, you will need to install the SVG2TikZ extension. This Provides a simple way to convert SVG files and inkscape projects to TikZ code.

Ipe

Ipe is a powerful graphics editor that incorporates LaTeX. It, like inkscape requires an extension(ipelets) to export TikZ code. The extension is called Ipe2TikZ , although it has been out of development for a while it still works well, there is a more recently active fork available here.

PGFPlotsEdt

PGFPlotsEdt is a tool that can be used to create TikZ code for PGFPlots. It is a web-based tool that allows you to create plots and then export the TikZ code. It is particularly useful for creating plots quickly and easily or if you are not comfortable writing TikZ code by hand. It additionally allows you to upload data files to create plots from.

As this is a web-based tool, just be cautious of the information you give it.

Alternatives

LaTeX Picture Environment

The standard LaTeX {picture} environment allows you to create simple graphics, but little more.

PSTricks

PSTricks package is powerful enough to create any kind of graphic, but it is not really portable. It has more limited driver compatibility than TikZ and is not as widely used.

Compared to TikZ, PSTricks has a similar support base. There are many nice extra packages for special purpose situations that have been contributed by users over the last decade. The TikZ syntax is more consistent than the pstricks syntax as TikZ was developed “in a more centralized manner” and also “with the shortcomings on pstricks in mind”.

METAPOST

METAPOST is a powerful alternative to TikZ. It used to be an external program, which entailed a bunch of problems, but it is now built in.

The main advantage of METAPOST is that it is a programming language, so you can do anything you want with it. The main disadvantage is that it is a programming language, so you have to learn it.

More information on it can be found in the METAPOST manual. Or additionally in the METAFUN manual from context and the creator of LuaLaTeX and context Hans Hagen. METAFUN manual

Asymptote

Asymptote is a powerful vector graphics language that is particularly well-suited for creating complex graphics. It is similar to METAPOST in that it is a programming language, but it additionally requires a separate program to compile the code.

gnuplot

Gnuplot is a powerful plotting program that can be used to create graphics in LaTeX. It is particularly useful for creating 2D and 3D plots, but it can also be used to create other types of graphics. It is not as powerful as TikZ or PGF, but it is easier to use and can be used to create complex graphics. However, it also requires a separate program.