| 1 |
% Copyright 2019 by Till Tantau |
| 2 |
% |
| 3 |
% This file may be distributed and/or modified |
| 4 |
% |
| 5 |
% 1. under the LaTeX Project Public License and/or |
| 6 |
% 2. under the GNU Free Documentation License. |
| 7 |
% |
| 8 |
% See the file doc/generic/pgf/licenses/LICENSE for more details. |
| 9 |
|
| 10 |
|
| 11 |
\section{Style Sheets and Legends} |
| 12 |
\label{section-dv-style-sheets} |
| 13 |
|
| 14 |
\subsection{Overview} |
| 15 |
|
| 16 |
In many data visualizations, different sets of data need to be visualized in a |
| 17 |
single visualization. For instance, in a plot there might be a line for the |
| 18 |
sine of~$x$ and another line for the cosine of~$x$; in another visualization |
| 19 |
there might be a set of points representing data from a first experiment and |
| 20 |
another set of points representing data from a second experiment; and so on. In |
| 21 |
order to indicate to which data set a data point belongs, one might plot the |
| 22 |
curve of the sine in, say, black, and the curve of the cosine in red; we might |
| 23 |
plot the data from the first experiment using stars and the data from the second |
| 24 |
experiment using circles; and so on. Finally, at some place in the |
| 25 |
visualization -- either inside the data or in a legend next to it -- the |
| 26 |
meaning of the colors or symbols need to be explained. |
| 27 |
|
| 28 |
Just as you would like \tikzname\ to map the data points automatically onto the |
| 29 |
axes, you will also typically wish \tikzname\ to choose for instance the |
| 30 |
coloring of the lines automatically for you. This is done using \emph{style |
| 31 |
sheets}. There are at least two good reasons why you should prefer style sheets |
| 32 |
over configuring the styling of each visualizer ``by hand'' using the |style| |
| 33 |
key: |
| 34 |
% |
| 35 |
\begin{enumerate} |
| 36 |
\item It is far more convenient to just say |style sheet=strong colors| |
| 37 |
than having to individually picking the different colors. |
| 38 |
\item The style sheets were chosen and constructed rather carefully. |
| 39 |
|
| 40 |
For instance, the |strong colors| style sheet does not pick colors like |
| 41 |
pure green or pure yellow, which have very low contrast with respect to |
| 42 |
a white background and which often lead to unintelligible graphics. |
| 43 |
Instead, opposing primary colors with maximum contrast on a white |
| 44 |
background were picked that are visually quite pleasing. |
| 45 |
|
| 46 |
Similarly, the different dashing style sheets are constructed in such a |
| 47 |
way that there are only few and small gaps in the dashing so that no |
| 48 |
data points get lost because the dashes are spaced too far apart. Also |
| 49 |
dashing patterns were chosen that have a maximum optical difference. |
| 50 |
|
| 51 |
As a final example, style sheets for plot marks are constructed in such |
| 52 |
a way that even when two plot marks lie directly on top of each other, |
| 53 |
they are still easily distinguishable. |
| 54 |
\end{enumerate} |
| 55 |
% |
| 56 |
The bottom line is that whenever possible, you should use one of the predefined |
| 57 |
style sheets rather than picking colors or dashings at random. |
| 58 |
|
| 59 |
|
| 60 |
\subsection{Concepts: Style Sheets} |
| 61 |
|
| 62 |
A \emph{style sheet} is a predefined list of styles such as a list of colors, a |
| 63 |
list of dashing pattern, a list of plot marks, or a combinations thereof. A |
| 64 |
style sheet can be \emph{attached} to a data point attribute. Then, the value |
| 65 |
of this attribute is used with data points to choose which style in the list |
| 66 |
should be chosen to visualize the data point. |
| 67 |
|
| 68 |
In most cases, there is just one attribute to which style sheets get attached: |
| 69 |
the |/data point/visualizer| attribute. The effect of attaching a style sheet |
| 70 |
to this attribute is that each visualizer is styled differently. |
| 71 |
|
| 72 |
For the following examples, let us first define a simple data set: |
| 73 |
% |
| 74 |
\begin{codeexample}[preamble={\usetikzlibrary{datavisualization.formats.functions}}] |
| 75 |
\tikz \datavisualization data group {function classes} = { |
| 76 |
data [set=log, format=function] { |
| 77 |
var x : interval [0.2:2.5]; |
| 78 |
func y = ln(\value x); |
| 79 |
} |
| 80 |
data [set=lin, format=function] { |
| 81 |
var x : interval [-2:2.5]; |
| 82 |
func y = 0.5*\value x; |
| 83 |
} |
| 84 |
data [set=squared, format=function] { |
| 85 |
var x : interval [-1.5:1.5]; |
| 86 |
func y = \value x*\value x; |
| 87 |
} |
| 88 |
data [set=exp, format=function] { |
| 89 |
var x : interval [-2.5:1]; |
| 90 |
func y = exp(\value x); |
| 91 |
} |
| 92 |
}; |
| 93 |
\end{codeexample} |
| 94 |
|
| 95 |
\begin{codeexample}[ |
| 96 |
width=6cm, |
| 97 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 98 |
pre={\tikz \datavisualization data group {function classes} = { |
| 99 |
data [set=log, format=function] { |
| 100 |
var x : interval [0.2:2.5]; |
| 101 |
func y = ln(\value x); |
| 102 |
} |
| 103 |
data [set=lin, format=function] { |
| 104 |
var x : interval [-2:2.5]; |
| 105 |
func y = 0.5*\value x; |
| 106 |
} |
| 107 |
data [set=squared, format=function] { |
| 108 |
var x : interval [-1.5:1.5]; |
| 109 |
func y = \value x*\value x; |
| 110 |
} |
| 111 |
data [set=exp, format=function] { |
| 112 |
var x : interval [-2.5:1]; |
| 113 |
func y = exp(\value x); |
| 114 |
} |
| 115 |
};}] |
| 116 |
\tikz \datavisualization [ |
| 117 |
school book axes, all axes={unit length=7.5mm}, |
| 118 |
visualize as smooth line/.list={log, lin, squared, exp}, |
| 119 |
style sheet=strong colors] |
| 120 |
data group {function classes}; |
| 121 |
\end{codeexample} |
| 122 |
|
| 123 |
\begin{codeexample}[ |
| 124 |
width=6cm, |
| 125 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 126 |
pre={\tikz \datavisualization data group {function classes} = { |
| 127 |
data [set=log, format=function] { |
| 128 |
var x : interval [0.2:2.5]; |
| 129 |
func y = ln(\value x); |
| 130 |
} |
| 131 |
data [set=lin, format=function] { |
| 132 |
var x : interval [-2:2.5]; |
| 133 |
func y = 0.5*\value x; |
| 134 |
} |
| 135 |
data [set=squared, format=function] { |
| 136 |
var x : interval [-1.5:1.5]; |
| 137 |
func y = \value x*\value x; |
| 138 |
} |
| 139 |
data [set=exp, format=function] { |
| 140 |
var x : interval [-2.5:1]; |
| 141 |
func y = exp(\value x); |
| 142 |
} |
| 143 |
};}] |
| 144 |
\tikz \datavisualization [ |
| 145 |
school book axes, all axes={unit length=7.5mm}, |
| 146 |
visualize as smooth line/.list={log, lin, squared, exp}, |
| 147 |
style sheet=vary dashing] |
| 148 |
data group {function classes}; |
| 149 |
\end{codeexample} |
| 150 |
|
| 151 |
|
| 152 |
\subsection{Concepts: Legends} |
| 153 |
\label{section-dv-labels-in} |
| 154 |
|
| 155 |
A \emph{legend} is a box that is next to a data visualization (or inside it at |
| 156 |
some otherwise empty position) that contains a textual explanation of the |
| 157 |
different colors or styles used in a data visualization. |
| 158 |
|
| 159 |
Just as it is difficult to get colors and dashing patterns right ``by hand'', |
| 160 |
it is also difficult to get a legend right. For instance, when a small line is |
| 161 |
shown in the legend that represents the actual line in the data visualization, |
| 162 |
if the line is too short and the dashing is too large, it may be impossible to |
| 163 |
discern which dashing is actually meant. Similarly, when plot marks are shown |
| 164 |
on such a short line, using a simple straight line may make it hard to read the |
| 165 |
plot marks correctly. |
| 166 |
|
| 167 |
The data visualization engine makes some effort to make it easy to create |
| 168 |
high-quality legends. Additionally, it also offers ways of easily adding labels |
| 169 |
for visualizers directly inside the data visualization, which is even better |
| 170 |
than adding a legend, in general. |
| 171 |
% |
| 172 |
\begin{codeexample}[ |
| 173 |
width=7cm, |
| 174 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 175 |
pre={\tikz \datavisualization data group {function classes} = { |
| 176 |
data [set=log, format=function] { |
| 177 |
var x : interval [0.2:2.5]; |
| 178 |
func y = ln(\value x); |
| 179 |
} |
| 180 |
data [set=lin, format=function] { |
| 181 |
var x : interval [-2:2.5]; |
| 182 |
func y = 0.5*\value x; |
| 183 |
} |
| 184 |
data [set=squared, format=function] { |
| 185 |
var x : interval [-1.5:1.5]; |
| 186 |
func y = \value x*\value x; |
| 187 |
} |
| 188 |
data [set=exp, format=function] { |
| 189 |
var x : interval [-2.5:1]; |
| 190 |
func y = exp(\value x); |
| 191 |
} |
| 192 |
};}] |
| 193 |
\tikz \datavisualization [ |
| 194 |
school book axes, all axes={unit length=7.5mm}, |
| 195 |
x axis={label=$x$}, |
| 196 |
visualize as smooth line/.list={log, lin, squared, exp}, |
| 197 |
log= {label in legend={text=$\log x$}}, |
| 198 |
lin= {label in legend={text=$x/2$}}, |
| 199 |
squared={label in legend={text=$x^2$}}, |
| 200 |
exp= {label in legend={text=$e^x$}}, |
| 201 |
style sheet=vary dashing] |
| 202 |
data group {function classes}; |
| 203 |
\end{codeexample} |
| 204 |
|
| 205 |
\begin{codeexample}[ |
| 206 |
width=6.3cm, |
| 207 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 208 |
pre={\tikz \datavisualization data group {function classes} = { |
| 209 |
data [set=log, format=function] { |
| 210 |
var x : interval [0.2:2.5]; |
| 211 |
func y = ln(\value x); |
| 212 |
} |
| 213 |
data [set=lin, format=function] { |
| 214 |
var x : interval [-2:2.5]; |
| 215 |
func y = 0.5*\value x; |
| 216 |
} |
| 217 |
data [set=squared, format=function] { |
| 218 |
var x : interval [-1.5:1.5]; |
| 219 |
func y = \value x*\value x; |
| 220 |
} |
| 221 |
data [set=exp, format=function] { |
| 222 |
var x : interval [-2.5:1]; |
| 223 |
func y = exp(\value x); |
| 224 |
} |
| 225 |
};}] |
| 226 |
\tikz \datavisualization [ |
| 227 |
school book axes, |
| 228 |
x axis={label=$x$}, |
| 229 |
visualize as smooth line/.list={log, lin, squared, exp}, |
| 230 |
every data set label/.append style={text colored}, |
| 231 |
log= {pin in data={text'=$\log x$, when=y is -1}}, |
| 232 |
lin= {pin in data={text=$x/2$, when=x is 2, |
| 233 |
pin length=1ex}}, |
| 234 |
squared={pin in data={text=$x^2$, when=x is 1.1, |
| 235 |
pin angle=230}}, |
| 236 |
exp= {label in data={text=$e^x$, when=x is -2}}, |
| 237 |
style sheet=vary hue] |
| 238 |
data group {function classes}; |
| 239 |
\end{codeexample} |
| 240 |
|
| 241 |
|
| 242 |
\subsection{Usage: Style Sheets} |
| 243 |
|
| 244 |
\subsubsection{Picking a Style Sheet} |
| 245 |
|
| 246 |
To use a style sheet, you need to \emph{attach} it to an attribute. You can |
| 247 |
attach multiple style sheets to an attribute and in this case all of these |
| 248 |
style sheets can influence the appearance of the data points. |
| 249 |
|
| 250 |
Most of the time, you will attach a style sheet to the |set| attribute. This |
| 251 |
has the effect that each different data set inside the same visualization is |
| 252 |
rendered in a different way. Since this use of style sheets is the most common, |
| 253 |
there is a special, easy-to-remember option for this: |
| 254 |
|
| 255 |
\begin{key}{/tikz/data visualization/style sheet=\meta{style sheet}} |
| 256 |
Adds the \meta{style sheet} to the list of style sheets attached to the |
| 257 |
|set| attribute. |
| 258 |
% |
| 259 |
\begin{codeexample}[ |
| 260 |
width=6cm, |
| 261 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 262 |
pre={\tikz \datavisualization data group {function classes} = { |
| 263 |
data [set=log, format=function] { |
| 264 |
var x : interval [0.2:2.5]; |
| 265 |
func y = ln(\value x); |
| 266 |
} |
| 267 |
data [set=lin, format=function] { |
| 268 |
var x : interval [-2:2.5]; |
| 269 |
func y = 0.5*\value x; |
| 270 |
} |
| 271 |
data [set=squared, format=function] { |
| 272 |
var x : interval [-1.5:1.5]; |
| 273 |
func y = \value x*\value x; |
| 274 |
} |
| 275 |
data [set=exp, format=function] { |
| 276 |
var x : interval [-2.5:1]; |
| 277 |
func y = exp(\value x); |
| 278 |
} |
| 279 |
};}, |
| 280 |
] |
| 281 |
\tikz \datavisualization [ |
| 282 |
school book axes, all axes={unit length=7.5mm}, |
| 283 |
visualize as smooth line/.list={log, lin, squared, exp}, |
| 284 |
style sheet=vary thickness and dashing, |
| 285 |
style sheet=vary hue] |
| 286 |
data group {function classes}; |
| 287 |
\end{codeexample} |
| 288 |
% |
| 289 |
\end{key} |
| 290 |
|
| 291 |
While the |style sheet| key will attach a style sheet only to the |set| |
| 292 |
attribute, the following key handler can be used to attach a style sheet to an |
| 293 |
arbitrary attribute: |
| 294 |
|
| 295 |
\begin{handler}{{.style sheet}=\meta{style sheet}} |
| 296 |
Inside a data visualization you can use this key handler together with an |
| 297 |
attribute, that is, with a key having the path prefix |/data point|. For |
| 298 |
instance, in order to attach the \meta{style sheet} |strong colors| to the |
| 299 |
attribute |set|, you could write |
| 300 |
% |
| 301 |
\begin{codeexample}[code only] |
| 302 |
/data point/set/.style sheet=strong colors |
| 303 |
\end{codeexample} |
| 304 |
% |
| 305 |
Indeed, the |style sheet| key is just a shorthand for the above. |
| 306 |
|
| 307 |
The effect of attaching a style sheet is the following: |
| 308 |
% |
| 309 |
\begin{itemize} |
| 310 |
\item A new object is created that will monitor the attribute. |
| 311 |
\item Each time a special \emph{styling key} is emitted by the data |
| 312 |
visualization engine, this object will inspect the current value of |
| 313 |
the attribute to which it is attached. |
| 314 |
\item Depending on this value, one of the styles stored in the style |
| 315 |
sheet is chosen (how this works, exactly, will be explained in a |
| 316 |
moment). |
| 317 |
\item The chosen style is then locally applied. |
| 318 |
\end{itemize} |
| 319 |
|
| 320 |
In reality, things are a bit more complicated: If the attribute of the data |
| 321 |
point happens to have a subkey named in the same way as the value, then the |
| 322 |
value of is this subkey is used instead of the value itself. This allows |
| 323 |
you to ``rename'' a value. |
| 324 |
|
| 325 |
In a sense, a style sheet behaves much like a visualizer (see |
| 326 |
Section~\ref{section-dv-visualizers}): In accordance with the value of a |
| 327 |
certain attribute, the appearance of data points change. However, there are |
| 328 |
a few differences: First, the styling of a data point needs to be triggered |
| 329 |
explicitly and this triggering is not necessarily done for each data point |
| 330 |
individually, but only for a whole visualizer. Second, styles can be |
| 331 |
computed even when no data point is present. This is useful for instance in |
| 332 |
a legend since, here, a visual representation of a visualizer needs to be |
| 333 |
created independently of the actual data points. |
| 334 |
\end{handler} |
| 335 |
|
| 336 |
|
| 337 |
\subsubsection{Creating a New Style Sheet} |
| 338 |
|
| 339 |
Creating a style sheet works as follows: For each possible value that an |
| 340 |
attribute can attain we must specify a style. This is done by creating a style |
| 341 |
key for each such possible value with a special path prefix and setting this |
| 342 |
style key to the desired value. The special path prefix is |
| 343 |
|/pgf/data visualization/style sheets| followed by the name of the style sheet. |
| 344 |
|
| 345 |
As an example, suppose we wish to create a style sheet |test| that makes styled |
| 346 |
data points |red| when the attribute has value |foo| and |green| when the |
| 347 |
attribute has value |bar| and |dashed, blue| when the attribute is |foobar|. We |
| 348 |
could then write |
| 349 |
% |
| 350 |
\begin{codeexample}[code only] |
| 351 |
/pgf/data visualization/style sheets/test/foo/.style={red}, |
| 352 |
/pgf/data visualization/style sheets/test/bar/.style={green}, |
| 353 |
/pgf/data visualization/style sheets/test/foobar/.style={dashed, blue}, |
| 354 |
\end{codeexample} |
| 355 |
|
| 356 |
We could then attach this style sheet to the attribute |code| as follows: |
| 357 |
% |
| 358 |
\begin{codeexample}[code only] |
| 359 |
/data point/code/.style sheet=test |
| 360 |
\end{codeexample} |
| 361 |
|
| 362 |
Then, when |/data point/code=foobar| holds when the styling signal is raised, |
| 363 |
the style |dashed, blue| will get executed. |
| 364 |
|
| 365 |
A natural question arises concerning the situation that the value of the |
| 366 |
attribute is not defined as a subkey of the style sheet. In this case, a |
| 367 |
special key gets executed: |
| 368 |
|
| 369 |
\begin{stylekey}{/pgf/data visualization/style sheets/\meta{style sheet}/default style=\meta{value}} |
| 370 |
This key gets during styling whenever |
| 371 |
|/pgf/data visualization/style sheet/|\meta{style sheet}|/|\meta{value} is |
| 372 |
not defined. |
| 373 |
\end{stylekey} |
| 374 |
|
| 375 |
Let us put all of this together in a real-life example. Suppose we wish to |
| 376 |
create a style sheet that makes the first data set |green|, the second |yellow| |
| 377 |
and the third one |red|. Further data sets should be, say, |black|. The |
| 378 |
attribute that we intend to style is the |set| attribute. For the moment, we |
| 379 |
assume that the data sets will be named |1|, |2|, |3|, and so on (instead of, |
| 380 |
say, |experiment 1| or |sin| or something more readable -- we will get rid of |
| 381 |
this restriction in a minute). |
| 382 |
|
| 383 |
We would now write: |
| 384 |
% |
| 385 |
\begin{codeexample}[preamble={\usetikzlibrary{datavisualization}}] |
| 386 |
\pgfkeys{ |
| 387 |
/pgf/data visualization/style sheets/traffic light/.cd, |
| 388 |
% All these styles have the above prefix. |
| 389 |
1/.style={green!50!black}, |
| 390 |
2/.style={yellow!90!black}, |
| 391 |
3/.style={red!80!black}, |
| 392 |
default style/.style={black} |
| 393 |
} |
| 394 |
\tikz \datavisualization [ |
| 395 |
school book axes, |
| 396 |
visualize as line=1, |
| 397 |
visualize as line=2, |
| 398 |
visualize as line=3, |
| 399 |
style sheet=traffic light] |
| 400 |
data point [x=0, y=0, set=1] |
| 401 |
data point [x=2, y=2, set=1] |
| 402 |
data point [x=0, y=1, set=2] |
| 403 |
data point [x=2, y=1, set=2] |
| 404 |
data point [x=0.5, y=1.5, set=3] |
| 405 |
data point [x=2.25, y=1.75, set=3]; |
| 406 |
\end{codeexample} |
| 407 |
|
| 408 |
In the above example, we have to name the visualizers |1|, |2|, |3| and so one |
| 409 |
since the value of the |set| attribute is used both assign data points to |
| 410 |
visualizers and also pick a style sheet. However, it would be much nicer if we |
| 411 |
could name any way we want. To achieve this, we use the special rule for style |
| 412 |
sheets that says that if there is a subkey of an attribute whose name is the |
| 413 |
same name as the value, then the value of this key is used instead. This |
| 414 |
slightly intimidating definition is much easier to understand when we have a |
| 415 |
look at an example: |
| 416 |
% |
| 417 |
\pgfkeys{ |
| 418 |
/pgf/data visualization/style sheets/traffic light/.cd, |
| 419 |
% All these styles have the above prefix. |
| 420 |
1/.style={green!50!black}, |
| 421 |
2/.style={yellow!90!black}, |
| 422 |
3/.style={red!80!black}, |
| 423 |
default style/.style={black} |
| 424 |
} |
| 425 |
\begin{codeexample}[ |
| 426 |
preamble={\usetikzlibrary{datavisualization}}, |
| 427 |
pre={\pgfkeys{ |
| 428 |
/pgf/data visualization/style sheets/traffic light/.cd, |
| 429 |
% All these styles have the above prefix. |
| 430 |
1/.style={green!50!black}, |
| 431 |
2/.style={yellow!90!black}, |
| 432 |
3/.style={red!80!black}, |
| 433 |
default style/.style={black} |
| 434 |
}}, |
| 435 |
] |
| 436 |
% Definition of traffic light keys as above |
| 437 |
\begin{tikzpicture} |
| 438 |
\datavisualization data group {lines} = { |
| 439 |
data point [x=0, y=0, set=normal] |
| 440 |
data point [x=2, y=2, set=normal] |
| 441 |
data point [x=0, y=1, set=heated] |
| 442 |
data point [x=2, y=1, set=heated] |
| 443 |
data point [x=0.5, y=1.5, set=critical] |
| 444 |
data point [x=2.25, y=1.75, set=critical] |
| 445 |
}; |
| 446 |
\datavisualization [ |
| 447 |
school book axes, |
| 448 |
visualize as line=normal, |
| 449 |
visualize as line=heated, |
| 450 |
visualize as line=critical, |
| 451 |
/data point/set/normal/.initial=1, |
| 452 |
/data point/set/heated/.initial=2, |
| 453 |
/data point/set/critical/.initial=3, |
| 454 |
style sheet=traffic light] |
| 455 |
data group {lines}; |
| 456 |
\end{tikzpicture} |
| 457 |
\end{codeexample} |
| 458 |
|
| 459 |
Now, it is a bit bothersome that we have to set all these |/data point/set/...| |
| 460 |
keys by hand. It turns out that this is not necessary: Each time a visualizer |
| 461 |
is created, a subkey of |/data point/set| with the name of the visualizer is |
| 462 |
created automatically and a number is stored that is increased for each new |
| 463 |
visualizer in a data visualization. This means that the three lines starting |
| 464 |
with |/data point| are inserted automatically for you, so they can be left out. |
| 465 |
However, you would need them for instance when you would like several different |
| 466 |
data sets to use the same styling: |
| 467 |
% |
| 468 |
\begin{codeexample}[ |
| 469 |
preamble={\usetikzlibrary{datavisualization}}, |
| 470 |
pre={\pgfkeys{ |
| 471 |
/pgf/data visualization/style sheets/traffic light/.cd, |
| 472 |
% All these styles have the above prefix. |
| 473 |
1/.style={green!50!black}, |
| 474 |
2/.style={yellow!90!black}, |
| 475 |
3/.style={red!80!black}, |
| 476 |
default style/.style={black} |
| 477 |
}% |
| 478 |
\tikz \datavisualization data group {lines} = { |
| 479 |
data point [x=0, y=0, set=normal] |
| 480 |
data point [x=2, y=2, set=normal] |
| 481 |
data point [x=0, y=1, set=heated] |
| 482 |
data point [x=2, y=1, set=heated] |
| 483 |
data point [x=0.5, y=1.5, set=critical] |
| 484 |
data point [x=2.25, y=1.75, set=critical] |
| 485 |
};}, |
| 486 |
] |
| 487 |
% Definition of traffic light keys as above |
| 488 |
\tikz \datavisualization [ |
| 489 |
school book axes, |
| 490 |
visualize as line=normal, |
| 491 |
visualize as line=heated, |
| 492 |
visualize as line=critical, |
| 493 |
/data point/set/critical/.initial=1, % same styling as first set |
| 494 |
style sheet=traffic light] |
| 495 |
data group {lines}; |
| 496 |
\end{codeexample} |
| 497 |
|
| 498 |
We can a command that slightly simplifies the definition of style sheets: |
| 499 |
|
| 500 |
\begin{command}{\pgfdvdeclarestylesheet\marg{name}\marg{keys}} |
| 501 |
This command executes the \meta{keys} with the path prefix |
| 502 |
|/pgf/data visualization/style sheets/|\penalty0\meta{name}. The above |
| 503 |
definition of the traffic light style sheet could be rewritten as follows: |
| 504 |
% |
| 505 |
\begin{codeexample}[code only] |
| 506 |
\pgfdvdeclarestylesheet{traffic light}{ |
| 507 |
1/.style={green!50!black}, |
| 508 |
2/.style={yellow!90!black}, |
| 509 |
3/.style={red!80!black}, |
| 510 |
default style/.style={black} |
| 511 |
} |
| 512 |
\end{codeexample} |
| 513 |
% |
| 514 |
\end{command} |
| 515 |
|
| 516 |
As a final example, let us create a style sheet that changes the dashing |
| 517 |
pattern according to the value of the attribute. We do not need to define an |
| 518 |
large number of styles in this case, but can use the |default style| key to |
| 519 |
``calculate'' the correct dashing. |
| 520 |
|
| 521 |
\begin{codeexample}[ |
| 522 |
preamble={\usetikzlibrary{datavisualization}}, |
| 523 |
pre={\tikz \datavisualization data group {lines} = { |
| 524 |
data point [x=0, y=0, set=normal] |
| 525 |
data point [x=2, y=2, set=normal] |
| 526 |
data point [x=0, y=1, set=heated] |
| 527 |
data point [x=2, y=1, set=heated] |
| 528 |
data point [x=0.5, y=1.5, set=critical] |
| 529 |
data point [x=2.25, y=1.75, set=critical] |
| 530 |
};}, |
| 531 |
] |
| 532 |
\pgfdvdeclarestylesheet{my dashings}{ |
| 533 |
default style/.style={dash pattern={on #1pt off 1pt}} |
| 534 |
} |
| 535 |
\tikz \datavisualization [ |
| 536 |
school book axes, |
| 537 |
visualize as line=normal, |
| 538 |
visualize as line=heated, |
| 539 |
visualize as line=critical, |
| 540 |
style sheet=my dashings] |
| 541 |
data group {lines}; |
| 542 |
\end{codeexample} |
| 543 |
|
| 544 |
|
| 545 |
\subsubsection{Creating a New Color Style Sheet} |
| 546 |
|
| 547 |
Creating a style sheet that varies colors according to an attribute works the |
| 548 |
same way as creating a normal style sheet: Subkeys lies |1|, |2|, and so on use |
| 549 |
the |style| attribute to setup a color. However, instead of using the |color| |
| 550 |
attribute to set the color, you should use the |visualizer color| key to set |
| 551 |
the color: |
| 552 |
|
| 553 |
\begin{key}{/tikz/visualizer color=\meta{color}} |
| 554 |
This key is used to set the color |visualizer color| to \meta{color}. This |
| 555 |
color is used by visualizers to color the data they visualize, rather than |
| 556 |
the current ``standard color''. The reason for not using the normal current |
| 557 |
color is simply that it makes many internals of the data visualization |
| 558 |
engine a bit simpler. |
| 559 |
% |
| 560 |
\begin{codeexample}[ |
| 561 |
preamble={\usetikzlibrary{datavisualization}}, |
| 562 |
pre={\tikz \datavisualization data group {lines} = { |
| 563 |
data point [x=0, y=0, set=normal] |
| 564 |
data point [x=2, y=2, set=normal] |
| 565 |
data point [x=0, y=1, set=heated] |
| 566 |
data point [x=2, y=1, set=heated] |
| 567 |
data point [x=0.5, y=1.5, set=critical] |
| 568 |
data point [x=2.25, y=1.75, set=critical] |
| 569 |
};}, |
| 570 |
] |
| 571 |
\pgfdvdeclarestylesheet{my colors} |
| 572 |
{ |
| 573 |
default style/.style={visualizer color=black}, |
| 574 |
1/.style={visualizer color=black}, |
| 575 |
2/.style={visualizer color=red!80!black}, |
| 576 |
3/.style={visualizer color=blue!80!black}, |
| 577 |
} |
| 578 |
\tikz \datavisualization [ |
| 579 |
school book axes, |
| 580 |
visualize as line=normal, |
| 581 |
visualize as line=heated, |
| 582 |
visualize as line=critical, |
| 583 |
style sheet=my colors] |
| 584 |
data group {lines}; |
| 585 |
\end{codeexample} |
| 586 |
% |
| 587 |
\end{key} |
| 588 |
|
| 589 |
There is an additional command that makes it easy to define a style sheet based |
| 590 |
on a \emph{color series}. Color series are a concept from the |xcolor| package: |
| 591 |
The idea is that we start with a certain color for the first data set and then |
| 592 |
add a certain ``color offset'' for each next data point. Please consult the |
| 593 |
documentation of the |xcolor| package for details. |
| 594 |
|
| 595 |
\begin{command}{\tikzdvdeclarestylesheetcolorseries\marg{name}\marg{color model}\marg{initial color}\marg{step}} |
| 596 |
This command creates a new style sheet using |\pgfdvdeclarestylesheet|. |
| 597 |
This style sheet will only have a default style setup that maps numbers to |
| 598 |
the color in the color series starting with \meta{initial color} and having |
| 599 |
a stepping of \meta{step}. Note that when the value of the attribute is |
| 600 |
|1|, which it is the first data set, the \emph{second} color in the color |
| 601 |
series is used (since counting starts at |0| for color series). Thus, in |
| 602 |
general, you need to start the \meta{initial color} ``one early''. |
| 603 |
% |
| 604 |
\begin{codeexample}[ |
| 605 |
preamble={\usetikzlibrary{datavisualization}}, |
| 606 |
pre={\tikz \datavisualization data group {lines} = { |
| 607 |
data point [x=0, y=0, set=normal] |
| 608 |
data point [x=2, y=2, set=normal] |
| 609 |
data point [x=0, y=1, set=heated] |
| 610 |
data point [x=2, y=1, set=heated] |
| 611 |
data point [x=0.5, y=1.5, set=critical] |
| 612 |
data point [x=2.25, y=1.75, set=critical] |
| 613 |
};}, |
| 614 |
] |
| 615 |
\tikzdvdeclarestylesheetcolorseries{greens}{hsb}{0.3,1.3,0.8}{0,-.4,-.1} |
| 616 |
\tikz \datavisualization [ |
| 617 |
school book axes, |
| 618 |
visualize as line=normal, |
| 619 |
visualize as line=heated, |
| 620 |
visualize as line=critical, |
| 621 |
style sheet=greens] |
| 622 |
data group {lines}; |
| 623 |
\end{codeexample} |
| 624 |
% |
| 625 |
\end{command} |
| 626 |
|
| 627 |
|
| 628 |
\subsection{Reference: Style Sheets for Lines} |
| 629 |
|
| 630 |
The following style sheets can be applied to visualizations that use the |
| 631 |
|visualize as line| and related keys. For the examples, the following style and |
| 632 |
data set are used: |
| 633 |
% |
| 634 |
\begin{codeexample}[code only] |
| 635 |
\tikzdatavisualizationset { |
| 636 |
example visualization/.style={ |
| 637 |
scientific axes=clean, |
| 638 |
y axis={ticks={style={ |
| 639 |
/pgf/number format/fixed, |
| 640 |
/pgf/number format/fixed zerofill, |
| 641 |
/pgf/number format/precision=2}}}, |
| 642 |
x axis={ticks={tick suffix=${}^\circ$}}, |
| 643 |
1={label in legend={text=$\frac{1}{6}\sin 11x$}}, |
| 644 |
2={label in legend={text=$\frac{1}{7}\sin 12x$}}, |
| 645 |
3={label in legend={text=$\frac{1}{8}\sin 13x$}}, |
| 646 |
4={label in legend={text=$\frac{1}{9}\sin 14x$}}, |
| 647 |
5={label in legend={text=$\frac{1}{10}\sin 15x$}}, |
| 648 |
6={label in legend={text=$\frac{1}{11}\sin 16x$}}, |
| 649 |
7={label in legend={text=$\frac{1}{12}\sin 17x$}}, |
| 650 |
8={label in legend={text=$\frac{1}{13}\sin 18x$}} |
| 651 |
} |
| 652 |
} |
| 653 |
\end{codeexample} |
| 654 |
\tikzdatavisualizationset { |
| 655 |
example visualization/.style={ |
| 656 |
scientific axes=clean, |
| 657 |
y axis={ticks={style={ |
| 658 |
/pgf/number format/fixed, |
| 659 |
/pgf/number format/fixed zerofill, |
| 660 |
/pgf/number format/precision=2}}}, |
| 661 |
x axis={ticks={tick suffix=${}^\circ$}}, |
| 662 |
1={label in legend={text=$\frac{1}{6}\sin 11x$}}, |
| 663 |
2={label in legend={text=$\frac{1}{7}\sin 12x$}}, |
| 664 |
3={label in legend={text=$\frac{1}{8}\sin 13x$}}, |
| 665 |
4={label in legend={text=$\frac{1}{9}\sin 14x$}}, |
| 666 |
5={label in legend={text=$\frac{1}{10}\sin 15x$}}, |
| 667 |
6={label in legend={text=$\frac{1}{11}\sin 16x$}}, |
| 668 |
7={label in legend={text=$\frac{1}{12}\sin 17x$}}, |
| 669 |
8={label in legend={text=$\frac{1}{13}\sin 18x$}} |
| 670 |
} |
| 671 |
} |
| 672 |
|
| 673 |
\begin{codeexample}[code only] |
| 674 |
\tikz \datavisualization data group {sin functions} = { |
| 675 |
data [format=function] { |
| 676 |
var set : {1,...,8}; |
| 677 |
var x : interval [0:50]; |
| 678 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 679 |
} |
| 680 |
}; |
| 681 |
\end{codeexample} |
| 682 |
\tikz \datavisualization data group {sin functions} = { |
| 683 |
data [format=function] { |
| 684 |
var set : {1,...,8}; |
| 685 |
var x : interval [0:50]; |
| 686 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 687 |
} |
| 688 |
}; |
| 689 |
|
| 690 |
\begin{stylesheet}{vary thickness} |
| 691 |
This style varies the thickness of lines. It should be used only when there |
| 692 |
are only two or three lines, and even then it is not particularly pleasing |
| 693 |
visually. |
| 694 |
% |
| 695 |
\begin{codeexample}[ |
| 696 |
width=10cm, |
| 697 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 698 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 699 |
data [format=function] { |
| 700 |
var set : {1,...,8}; |
| 701 |
var x : interval [0:50]; |
| 702 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 703 |
} |
| 704 |
};% |
| 705 |
\tikzdatavisualizationset { |
| 706 |
example visualization/.style={ |
| 707 |
scientific axes=clean, |
| 708 |
y axis={ticks={style={ |
| 709 |
/pgf/number format/fixed, |
| 710 |
/pgf/number format/fixed zerofill, |
| 711 |
/pgf/number format/precision=2}}}, |
| 712 |
x axis={ticks={tick suffix=${}^\circ$}}, |
| 713 |
1={label in legend={text=$\frac{1}{6}\sin 11x$}}, |
| 714 |
2={label in legend={text=$\frac{1}{7}\sin 12x$}}, |
| 715 |
3={label in legend={text=$\frac{1}{8}\sin 13x$}}, |
| 716 |
4={label in legend={text=$\frac{1}{9}\sin 14x$}}, |
| 717 |
5={label in legend={text=$\frac{1}{10}\sin 15x$}}, |
| 718 |
6={label in legend={text=$\frac{1}{11}\sin 16x$}}, |
| 719 |
7={label in legend={text=$\frac{1}{12}\sin 17x$}}, |
| 720 |
8={label in legend={text=$\frac{1}{13}\sin 18x$}} |
| 721 |
} |
| 722 |
}}, |
| 723 |
] |
| 724 |
\tikz \datavisualization [ |
| 725 |
visualize as smooth line/.list= |
| 726 |
{1,2,3,4,5,6,7,8}, |
| 727 |
example visualization, |
| 728 |
style sheet=vary thickness] |
| 729 |
data group {sin functions}; |
| 730 |
\end{codeexample} |
| 731 |
% |
| 732 |
\end{stylesheet} |
| 733 |
|
| 734 |
\begin{stylesheet}{vary dashing} |
| 735 |
This style varies the dashing of lines. Although it is not particularly |
| 736 |
pleasing visually and although visualizations using this style sheet tend |
| 737 |
to look ``excited'' (but not necessarily ``exciting''), this style sheet is |
| 738 |
often the best choice when the visualization is to be printed in black and |
| 739 |
white. |
| 740 |
% |
| 741 |
\begin{codeexample}[ |
| 742 |
width=10cm, |
| 743 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 744 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 745 |
data [format=function] { |
| 746 |
var set : {1,...,8}; |
| 747 |
var x : interval [0:50]; |
| 748 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 749 |
} |
| 750 |
};% |
| 751 |
\tikzdatavisualizationset { |
| 752 |
example visualization/.style={ |
| 753 |
scientific axes=clean, |
| 754 |
y axis={ticks={style={ |
| 755 |
/pgf/number format/fixed, |
| 756 |
/pgf/number format/fixed zerofill, |
| 757 |
/pgf/number format/precision=2}}}, |
| 758 |
x axis={ticks={tick suffix=${}^\circ$}}, |
| 759 |
1={label in legend={text=$\frac{1}{6}\sin 11x$}}, |
| 760 |
2={label in legend={text=$\frac{1}{7}\sin 12x$}}, |
| 761 |
3={label in legend={text=$\frac{1}{8}\sin 13x$}}, |
| 762 |
4={label in legend={text=$\frac{1}{9}\sin 14x$}}, |
| 763 |
5={label in legend={text=$\frac{1}{10}\sin 15x$}}, |
| 764 |
6={label in legend={text=$\frac{1}{11}\sin 16x$}}, |
| 765 |
7={label in legend={text=$\frac{1}{12}\sin 17x$}}, |
| 766 |
8={label in legend={text=$\frac{1}{13}\sin 18x$}} |
| 767 |
} |
| 768 |
}}, |
| 769 |
] |
| 770 |
\tikz \datavisualization [ |
| 771 |
visualize as smooth line/.list= |
| 772 |
{1,2,3,4,5,6,7,8}, |
| 773 |
example visualization, |
| 774 |
style sheet=vary dashing] |
| 775 |
data group {sin functions}; |
| 776 |
\end{codeexample} |
| 777 |
% |
| 778 |
As can be seen, there are only seven distinct dashing patterns. The eighth |
| 779 |
and further lines will use a solid line once more. You will then have to |
| 780 |
specify the dashing ``by hand'' using the |style| option together with the |
| 781 |
visualizer. |
| 782 |
\end{stylesheet} |
| 783 |
|
| 784 |
\begin{stylesheet}{vary thickness and dashing} |
| 785 |
This style alternates between varying the thickness and the dashing of |
| 786 |
lines. The difference to just using both the |vary thickness| and |
| 787 |
|vary dashing| is that too thick lines are avoided. Instead, this style |
| 788 |
creates clearly distinguishable line styles for many lines (up to 14) with |
| 789 |
a minimum of visual clutter. This style is the most useful for |
| 790 |
visualizations when many different lines (ten or more) should be printed in |
| 791 |
black and white. |
| 792 |
% |
| 793 |
\begin{codeexample}[ |
| 794 |
width=10cm, |
| 795 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 796 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 797 |
data [format=function] { |
| 798 |
var set : {1,...,8}; |
| 799 |
var x : interval [0:50]; |
| 800 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 801 |
} |
| 802 |
};% |
| 803 |
\tikzdatavisualizationset { |
| 804 |
example visualization/.style={ |
| 805 |
scientific axes=clean, |
| 806 |
y axis={ticks={style={ |
| 807 |
/pgf/number format/fixed, |
| 808 |
/pgf/number format/fixed zerofill, |
| 809 |
/pgf/number format/precision=2}}}, |
| 810 |
x axis={ticks={tick suffix=${}^\circ$}}, |
| 811 |
1={label in legend={text=$\frac{1}{6}\sin 11x$}}, |
| 812 |
2={label in legend={text=$\frac{1}{7}\sin 12x$}}, |
| 813 |
3={label in legend={text=$\frac{1}{8}\sin 13x$}}, |
| 814 |
4={label in legend={text=$\frac{1}{9}\sin 14x$}}, |
| 815 |
5={label in legend={text=$\frac{1}{10}\sin 15x$}}, |
| 816 |
6={label in legend={text=$\frac{1}{11}\sin 16x$}}, |
| 817 |
7={label in legend={text=$\frac{1}{12}\sin 17x$}}, |
| 818 |
8={label in legend={text=$\frac{1}{13}\sin 18x$}} |
| 819 |
} |
| 820 |
}}, |
| 821 |
] |
| 822 |
\tikz \datavisualization [ |
| 823 |
visualize as smooth line/.list= |
| 824 |
{1,2,3,4,5,6,7,8}, |
| 825 |
example visualization, |
| 826 |
style sheet=vary thickness |
| 827 |
and dashing] |
| 828 |
data group {sin functions}; |
| 829 |
\end{codeexample} |
| 830 |
% |
| 831 |
For comparison, here is the must-less-than-satisfactory result of combining |
| 832 |
the two independent style sheets: |
| 833 |
% |
| 834 |
\begin{codeexample}[ |
| 835 |
width=10cm, |
| 836 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 837 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 838 |
data [format=function] { |
| 839 |
var set : {1,...,8}; |
| 840 |
var x : interval [0:50]; |
| 841 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 842 |
} |
| 843 |
};% |
| 844 |
\tikzdatavisualizationset { |
| 845 |
example visualization/.style={ |
| 846 |
scientific axes=clean, |
| 847 |
y axis={ticks={style={ |
| 848 |
/pgf/number format/fixed, |
| 849 |
/pgf/number format/fixed zerofill, |
| 850 |
/pgf/number format/precision=2}}}, |
| 851 |
x axis={ticks={tick suffix=${}^\circ$}}, |
| 852 |
1={label in legend={text=$\frac{1}{6}\sin 11x$}}, |
| 853 |
2={label in legend={text=$\frac{1}{7}\sin 12x$}}, |
| 854 |
3={label in legend={text=$\frac{1}{8}\sin 13x$}}, |
| 855 |
4={label in legend={text=$\frac{1}{9}\sin 14x$}}, |
| 856 |
5={label in legend={text=$\frac{1}{10}\sin 15x$}}, |
| 857 |
6={label in legend={text=$\frac{1}{11}\sin 16x$}}, |
| 858 |
7={label in legend={text=$\frac{1}{12}\sin 17x$}}, |
| 859 |
8={label in legend={text=$\frac{1}{13}\sin 18x$}} |
| 860 |
} |
| 861 |
}}, |
| 862 |
] |
| 863 |
\tikz \datavisualization [ |
| 864 |
visualize as smooth line/.list= |
| 865 |
{1,2,3,4,5,6,7,8}, |
| 866 |
example visualization, |
| 867 |
style sheet=vary thickness, |
| 868 |
style sheet=vary dashing] |
| 869 |
data group {sin functions}; |
| 870 |
\end{codeexample} |
| 871 |
% |
| 872 |
\end{stylesheet} |
| 873 |
|
| 874 |
|
| 875 |
\subsection{Reference: Style Sheets for Scatter Plots} |
| 876 |
|
| 877 |
The following style sheets can be used both for scatter plots and also with |
| 878 |
lines. In the latter case, the marks are added to the lines. |
| 879 |
|
| 880 |
\begin{stylesheet}{cross marks} |
| 881 |
This style uses different crosses to distinguish between the data points of |
| 882 |
different data sets. The crosses were chosen in such a way that when two |
| 883 |
different cross marks lie at the same coordinate, their overall shape |
| 884 |
allows one to still uniquely determine which marks are on top of each |
| 885 |
other. |
| 886 |
|
| 887 |
This style supports only up to six different data sets and requires the |
| 888 |
|plotmarks| library. |
| 889 |
% |
| 890 |
\begin{codeexample}[ |
| 891 |
width=10cm, |
| 892 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 893 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 894 |
data [format=function] { |
| 895 |
var set : {1,...,8}; |
| 896 |
var x : interval [0:50]; |
| 897 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 898 |
} |
| 899 |
};% |
| 900 |
\tikzdatavisualizationset { |
| 901 |
example visualization/.style={ |
| 902 |
scientific axes=clean, |
| 903 |
y axis={ticks={style={ |
| 904 |
/pgf/number format/fixed, |
| 905 |
/pgf/number format/fixed zerofill, |
| 906 |
/pgf/number format/precision=2}}}, |
| 907 |
x axis={ticks={tick suffix=${}^\circ$}}, |
| 908 |
1={label in legend={text=$\frac{1}{6}\sin 11x$}}, |
| 909 |
2={label in legend={text=$\frac{1}{7}\sin 12x$}}, |
| 910 |
3={label in legend={text=$\frac{1}{8}\sin 13x$}}, |
| 911 |
4={label in legend={text=$\frac{1}{9}\sin 14x$}}, |
| 912 |
5={label in legend={text=$\frac{1}{10}\sin 15x$}}, |
| 913 |
6={label in legend={text=$\frac{1}{11}\sin 16x$}}, |
| 914 |
7={label in legend={text=$\frac{1}{12}\sin 17x$}}, |
| 915 |
8={label in legend={text=$\frac{1}{13}\sin 18x$}} |
| 916 |
} |
| 917 |
}}, |
| 918 |
] |
| 919 |
\tikz \datavisualization [ |
| 920 |
visualize as scatter/.list= |
| 921 |
{1,2,3,4,5,6,7,8}, |
| 922 |
example visualization, |
| 923 |
style sheet=cross marks] |
| 924 |
data group {sin functions}; |
| 925 |
\end{codeexample} |
| 926 |
% |
| 927 |
\begin{codeexample}[ |
| 928 |
width=10cm, |
| 929 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 930 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 931 |
data [format=function] { |
| 932 |
var set : {1,...,8}; |
| 933 |
var x : interval [0:50]; |
| 934 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 935 |
} |
| 936 |
};% |
| 937 |
\tikzdatavisualizationset { |
| 938 |
example visualization/.style={ |
| 939 |
scientific axes=clean, |
| 940 |
y axis={ticks={style={ |
| 941 |
/pgf/number format/fixed, |
| 942 |
/pgf/number format/fixed zerofill, |
| 943 |
/pgf/number format/precision=2}}}, |
| 944 |
x axis={ticks={tick suffix=${}^\circ$}}, |
| 945 |
1={label in legend={text=$\frac{1}{6}\sin 11x$}}, |
| 946 |
2={label in legend={text=$\frac{1}{7}\sin 12x$}}, |
| 947 |
3={label in legend={text=$\frac{1}{8}\sin 13x$}}, |
| 948 |
4={label in legend={text=$\frac{1}{9}\sin 14x$}}, |
| 949 |
5={label in legend={text=$\frac{1}{10}\sin 15x$}}, |
| 950 |
6={label in legend={text=$\frac{1}{11}\sin 16x$}}, |
| 951 |
7={label in legend={text=$\frac{1}{12}\sin 17x$}}, |
| 952 |
8={label in legend={text=$\frac{1}{13}\sin 18x$}} |
| 953 |
} |
| 954 |
}}, |
| 955 |
] |
| 956 |
\tikz \datavisualization [ |
| 957 |
visualize as smooth line/.list= |
| 958 |
{1,2,3,4,5,6,7,8}, |
| 959 |
example visualization, |
| 960 |
style sheet=cross marks] |
| 961 |
data group {sin functions}; |
| 962 |
\end{codeexample} |
| 963 |
% |
| 964 |
\end{stylesheet} |
| 965 |
|
| 966 |
|
| 967 |
\subsection{Reference: Color Style Sheets} |
| 968 |
|
| 969 |
Color style sheets are very useful for creating visually pleasing data |
| 970 |
visualizations that contain multiple data sets. However, there are two things |
| 971 |
to keep in mind: |
| 972 |
% |
| 973 |
\begin{itemize} |
| 974 |
\item At some point, every data visualization is printed or photo copied in |
| 975 |
black and white by someone. In this case, data sets can often no longer |
| 976 |
be distinguished. |
| 977 |
\item A few people are color blind. They will not be able to distinguish |
| 978 |
between red and green lines (and some people are not even able to |
| 979 |
distinguish colors at all). |
| 980 |
\end{itemize} |
| 981 |
|
| 982 |
For these reasons, if there is any chance that the data visualization will be |
| 983 |
printed in black and white at some point, consider combining color style sheets |
| 984 |
with style sheets like |vary dashing| to make data sets distinguishable in all |
| 985 |
situations. |
| 986 |
|
| 987 |
\begin{stylesheet}{strong colors} |
| 988 |
This style sheets uses pure primary colors that can very easily be |
| 989 |
distinguished. Although not as visually pleasing as the |vary hue| style |
| 990 |
sheet, the visualizations are easier to read when this style sheet is used. |
| 991 |
Up to six different data sets are supported. |
| 992 |
% |
| 993 |
\begin{codeexample}[ |
| 994 |
width=10cm, |
| 995 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 996 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 997 |
data [format=function] { |
| 998 |
var set : {1,...,8}; |
| 999 |
var x : interval [0:50]; |
| 1000 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 1001 |
} |
| 1002 |
};% |
| 1003 |
\tikzdatavisualizationset { |
| 1004 |
example visualization/.style={ |
| 1005 |
scientific axes=clean, |
| 1006 |
y axis={ticks={style={ |
| 1007 |
/pgf/number format/fixed, |
| 1008 |
/pgf/number format/fixed zerofill, |
| 1009 |
/pgf/number format/precision=2}}}, |
| 1010 |
x axis={ticks={tick suffix=${}^\circ$}}, |
| 1011 |
1={label in legend={text=$\frac{1}{6}\sin 11x$}}, |
| 1012 |
2={label in legend={text=$\frac{1}{7}\sin 12x$}}, |
| 1013 |
3={label in legend={text=$\frac{1}{8}\sin 13x$}}, |
| 1014 |
4={label in legend={text=$\frac{1}{9}\sin 14x$}}, |
| 1015 |
5={label in legend={text=$\frac{1}{10}\sin 15x$}}, |
| 1016 |
6={label in legend={text=$\frac{1}{11}\sin 16x$}}, |
| 1017 |
7={label in legend={text=$\frac{1}{12}\sin 17x$}}, |
| 1018 |
8={label in legend={text=$\frac{1}{13}\sin 18x$}} |
| 1019 |
} |
| 1020 |
}}, |
| 1021 |
] |
| 1022 |
\tikz \datavisualization [ |
| 1023 |
visualize as smooth line/.list= |
| 1024 |
{1,2,3,4,5,6,7,8}, |
| 1025 |
example visualization, |
| 1026 |
style sheet=strong colors] |
| 1027 |
data group {sin functions}; |
| 1028 |
\end{codeexample} |
| 1029 |
% |
| 1030 |
\begin{codeexample}[ |
| 1031 |
width=10cm, |
| 1032 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 1033 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 1034 |
data [format=function] { |
| 1035 |
var set : {1,...,8}; |
| 1036 |
var x : interval [0:50]; |
| 1037 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 1038 |
} |
| 1039 |
};% |
| 1040 |
\tikzdatavisualizationset { |
| 1041 |
example visualization/.style={ |
| 1042 |
scientific axes=clean, |
| 1043 |
y axis={ticks={style={ |
| 1044 |
/pgf/number format/fixed, |
| 1045 |
/pgf/number format/fixed zerofill, |
| 1046 |
/pgf/number format/precision=2}}}, |
| 1047 |
x axis={ticks={tick suffix=${}^\circ$}}, |
| 1048 |
1={label in legend={text=$\frac{1}{6}\sin 11x$}}, |
| 1049 |
2={label in legend={text=$\frac{1}{7}\sin 12x$}}, |
| 1050 |
3={label in legend={text=$\frac{1}{8}\sin 13x$}}, |
| 1051 |
4={label in legend={text=$\frac{1}{9}\sin 14x$}}, |
| 1052 |
5={label in legend={text=$\frac{1}{10}\sin 15x$}}, |
| 1053 |
6={label in legend={text=$\frac{1}{11}\sin 16x$}}, |
| 1054 |
7={label in legend={text=$\frac{1}{12}\sin 17x$}}, |
| 1055 |
8={label in legend={text=$\frac{1}{13}\sin 18x$}} |
| 1056 |
} |
| 1057 |
}}, |
| 1058 |
] |
| 1059 |
\tikz \datavisualization [ |
| 1060 |
visualize as smooth line/.list= |
| 1061 |
{1,2,3,4,5,6,7,8}, |
| 1062 |
example visualization, |
| 1063 |
style sheet=strong colors, |
| 1064 |
style sheet=vary dashing] |
| 1065 |
data group {sin functions}; |
| 1066 |
\end{codeexample} |
| 1067 |
% |
| 1068 |
\end{stylesheet} |
| 1069 |
|
| 1070 |
|
| 1071 |
Unlike |strong colors|, the following style sheets support, in principle, an |
| 1072 |
unlimited number of data set. In practice, as always, more than four or five |
| 1073 |
data sets lead to nearly indistinguishable data sets. |
| 1074 |
|
| 1075 |
\begin{stylesheet}{vary hue} |
| 1076 |
This style uses a different hue for each data set. |
| 1077 |
% |
| 1078 |
\begin{codeexample}[ |
| 1079 |
width=10cm, |
| 1080 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 1081 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 1082 |
data [format=function] { |
| 1083 |
var set : {1,...,8}; |
| 1084 |
var x : interval [0:50]; |
| 1085 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 1086 |
} |
| 1087 |
};% |
| 1088 |
\tikzdatavisualizationset { |
| 1089 |
example visualization/.style={ |
| 1090 |
scientific axes=clean, |
| 1091 |
y axis={ticks={style={ |
| 1092 |
/pgf/number format/fixed, |
| 1093 |
/pgf/number format/fixed zerofill, |
| 1094 |
/pgf/number format/precision=2}}}, |
| 1095 |
x axis={ticks={tick suffix=${}^\circ$}}, |
| 1096 |
1={label in legend={text=$\frac{1}{6}\sin 11x$}}, |
| 1097 |
2={label in legend={text=$\frac{1}{7}\sin 12x$}}, |
| 1098 |
3={label in legend={text=$\frac{1}{8}\sin 13x$}}, |
| 1099 |
4={label in legend={text=$\frac{1}{9}\sin 14x$}}, |
| 1100 |
5={label in legend={text=$\frac{1}{10}\sin 15x$}}, |
| 1101 |
6={label in legend={text=$\frac{1}{11}\sin 16x$}}, |
| 1102 |
7={label in legend={text=$\frac{1}{12}\sin 17x$}}, |
| 1103 |
8={label in legend={text=$\frac{1}{13}\sin 18x$}} |
| 1104 |
} |
| 1105 |
}}, |
| 1106 |
] |
| 1107 |
\tikz \datavisualization [ |
| 1108 |
visualize as smooth line/.list= |
| 1109 |
{1,2,3,4,5,6,7,8}, |
| 1110 |
example visualization, |
| 1111 |
style sheet=vary hue] |
| 1112 |
data group {sin functions}; |
| 1113 |
\end{codeexample} |
| 1114 |
% |
| 1115 |
\end{stylesheet} |
| 1116 |
|
| 1117 |
\begin{stylesheet}{shades of blue} |
| 1118 |
As the name suggests, different shades of blue are used for different data |
| 1119 |
sets. |
| 1120 |
\begin{codeexample}[ |
| 1121 |
width=10cm, |
| 1122 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 1123 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 1124 |
data [format=function] { |
| 1125 |
var set : {1,...,8}; |
| 1126 |
var x : interval [0:50]; |
| 1127 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 1128 |
} |
| 1129 |
};% |
| 1130 |
\tikzdatavisualizationset { |
| 1131 |
example visualization/.style={ |
| 1132 |
scientific axes=clean, |
| 1133 |
y axis={ticks={style={ |
| 1134 |
/pgf/number format/fixed, |
| 1135 |
/pgf/number format/fixed zerofill, |
| 1136 |
/pgf/number format/precision=2}}}, |
| 1137 |
x axis={ticks={tick suffix=${}^\circ$}}, |
| 1138 |
1={label in legend={text=$\frac{1}{6}\sin 11x$}}, |
| 1139 |
2={label in legend={text=$\frac{1}{7}\sin 12x$}}, |
| 1140 |
3={label in legend={text=$\frac{1}{8}\sin 13x$}}, |
| 1141 |
4={label in legend={text=$\frac{1}{9}\sin 14x$}}, |
| 1142 |
5={label in legend={text=$\frac{1}{10}\sin 15x$}}, |
| 1143 |
6={label in legend={text=$\frac{1}{11}\sin 16x$}}, |
| 1144 |
7={label in legend={text=$\frac{1}{12}\sin 17x$}}, |
| 1145 |
8={label in legend={text=$\frac{1}{13}\sin 18x$}} |
| 1146 |
} |
| 1147 |
}}, |
| 1148 |
] |
| 1149 |
\tikz \datavisualization [ |
| 1150 |
visualize as smooth line/.list= |
| 1151 |
{1,2,3,4,5,6,7,8}, |
| 1152 |
example visualization, |
| 1153 |
style sheet=shades of blue] |
| 1154 |
data group {sin functions}; |
| 1155 |
\end{codeexample} |
| 1156 |
% |
| 1157 |
\end{stylesheet} |
| 1158 |
|
| 1159 |
\begin{stylesheet}{shades of red} |
| 1160 |
\begin{codeexample}[ |
| 1161 |
width=10cm, |
| 1162 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 1163 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 1164 |
data [format=function] { |
| 1165 |
var set : {1,...,8}; |
| 1166 |
var x : interval [0:50]; |
| 1167 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 1168 |
} |
| 1169 |
};% |
| 1170 |
\tikzdatavisualizationset { |
| 1171 |
example visualization/.style={ |
| 1172 |
scientific axes=clean, |
| 1173 |
y axis={ticks={style={ |
| 1174 |
/pgf/number format/fixed, |
| 1175 |
/pgf/number format/fixed zerofill, |
| 1176 |
/pgf/number format/precision=2}}}, |
| 1177 |
x axis={ticks={tick suffix=${}^\circ$}}, |
| 1178 |
1={label in legend={text=$\frac{1}{6}\sin 11x$}}, |
| 1179 |
2={label in legend={text=$\frac{1}{7}\sin 12x$}}, |
| 1180 |
3={label in legend={text=$\frac{1}{8}\sin 13x$}}, |
| 1181 |
4={label in legend={text=$\frac{1}{9}\sin 14x$}}, |
| 1182 |
5={label in legend={text=$\frac{1}{10}\sin 15x$}}, |
| 1183 |
6={label in legend={text=$\frac{1}{11}\sin 16x$}}, |
| 1184 |
7={label in legend={text=$\frac{1}{12}\sin 17x$}}, |
| 1185 |
8={label in legend={text=$\frac{1}{13}\sin 18x$}} |
| 1186 |
} |
| 1187 |
}}, |
| 1188 |
] |
| 1189 |
\tikz \datavisualization [ |
| 1190 |
visualize as smooth line/.list= |
| 1191 |
{1,2,3,4,5,6,7,8}, |
| 1192 |
example visualization, |
| 1193 |
style sheet=shades of red] |
| 1194 |
data group {sin functions}; |
| 1195 |
\end{codeexample} |
| 1196 |
\end{stylesheet} |
| 1197 |
|
| 1198 |
\begin{stylesheet}{gray scale} |
| 1199 |
For once, this style sheet can also be used when the visualization is |
| 1200 |
printed in black and white. |
| 1201 |
% |
| 1202 |
\begin{codeexample}[ |
| 1203 |
width=10cm, |
| 1204 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 1205 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 1206 |
data [format=function] { |
| 1207 |
var set : {1,...,8}; |
| 1208 |
var x : interval [0:50]; |
| 1209 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 1210 |
} |
| 1211 |
};% |
| 1212 |
\tikzdatavisualizationset { |
| 1213 |
example visualization/.style={ |
| 1214 |
scientific axes=clean, |
| 1215 |
y axis={ticks={style={ |
| 1216 |
/pgf/number format/fixed, |
| 1217 |
/pgf/number format/fixed zerofill, |
| 1218 |
/pgf/number format/precision=2}}}, |
| 1219 |
x axis={ticks={tick suffix=${}^\circ$}}, |
| 1220 |
1={label in legend={text=$\frac{1}{6}\sin 11x$}}, |
| 1221 |
2={label in legend={text=$\frac{1}{7}\sin 12x$}}, |
| 1222 |
3={label in legend={text=$\frac{1}{8}\sin 13x$}}, |
| 1223 |
4={label in legend={text=$\frac{1}{9}\sin 14x$}}, |
| 1224 |
5={label in legend={text=$\frac{1}{10}\sin 15x$}}, |
| 1225 |
6={label in legend={text=$\frac{1}{11}\sin 16x$}}, |
| 1226 |
7={label in legend={text=$\frac{1}{12}\sin 17x$}}, |
| 1227 |
8={label in legend={text=$\frac{1}{13}\sin 18x$}} |
| 1228 |
} |
| 1229 |
}}, |
| 1230 |
] |
| 1231 |
\tikz \datavisualization [ |
| 1232 |
visualize as smooth line/.list= |
| 1233 |
{1,2,3,4,5,6,7,8}, |
| 1234 |
example visualization, |
| 1235 |
style sheet=gray scale] |
| 1236 |
data group {sin functions}; |
| 1237 |
\end{codeexample} |
| 1238 |
% |
| 1239 |
\end{stylesheet} |
| 1240 |
|
| 1241 |
|
| 1242 |
\subsection{Usage: Labeling Data Sets Inside the Visualization} |
| 1243 |
|
| 1244 |
In a visualization that contains multiple data sets, it is often necessary to |
| 1245 |
clearly point out which line or mark type corresponds to which data set. This |
| 1246 |
can be done in the main text via a sentence like ``the normal data (black) lies |
| 1247 |
clearly below the critical values (red)'', but it often a good idea to indicate |
| 1248 |
data sets ideally directly inside the data visualization or directly next to it |
| 1249 |
in a so-called legend. |
| 1250 |
|
| 1251 |
The data visualization engine has direct support both for indicating data sets |
| 1252 |
directly inside the visualization and also for indicating them in a legend. |
| 1253 |
|
| 1254 |
The ``best'' way of indicating where a data set lies or which color is used for |
| 1255 |
it is to put a label directly inside the data visualization. The reason this is |
| 1256 |
the ``best'' way is that people do not have to match the legend entries against |
| 1257 |
the data, let alone having to look up the meaning of line styles somewhere in |
| 1258 |
the text. However, adding a label directly inside the visualization is also the |
| 1259 |
most tricky way of indicating data sets since it is hard to compute good |
| 1260 |
positions for the labels automatically and since there needs to be some empty |
| 1261 |
space where the label can be put. |
| 1262 |
|
| 1263 |
|
| 1264 |
\subsubsection{Placing a Label Next to a Data Set} |
| 1265 |
|
| 1266 |
The following key is used to create a label inside the data visualization for a |
| 1267 |
data set: |
| 1268 |
|
| 1269 |
\begin{key}{/tikz/data visualization/visualizer options/label in data=\meta{options}} |
| 1270 |
This key is passed to a visualizer that has previously been created using |
| 1271 |
keys starting |visualize as ...|. It will create a label inside the data |
| 1272 |
visualization ``next'' to the visualizer (the details are explained in a |
| 1273 |
moment). You can use this key multiple times with a visualizer to create |
| 1274 |
multiple labels at different points with different texts. |
| 1275 |
|
| 1276 |
The \meta{options} determine which text is shown and where it is shown. |
| 1277 |
They are executed with the following path prefix: |
| 1278 |
% |
| 1279 |
\begin{codeexample}[code only] |
| 1280 |
/tikz/data visualization/visualizer label options |
| 1281 |
\end{codeexample} |
| 1282 |
|
| 1283 |
In order to configure which text is shown and where, use the following keys |
| 1284 |
inside the \meta{options}: |
| 1285 |
|
| 1286 |
\begin{key}{/tikz/data visualization/visualizer label options/text=\meta{text}} |
| 1287 |
This is the text that will be displayed next to the data. It will be to |
| 1288 |
the ``left'' of the data, see the description below. |
| 1289 |
\end{key} |
| 1290 |
% |
| 1291 |
\begin{key}{/tikz/data visualization/visualizer label options/text'=\meta{text}} |
| 1292 |
Like |text|, only the text will be to the ``right'' of the data. |
| 1293 |
\end{key} |
| 1294 |
|
| 1295 |
The following keys are used to configure where the label will be shown. |
| 1296 |
They use different strategies to specify one data point where the label |
| 1297 |
will be anchored. The coordinate of this data point will be stored in |
| 1298 |
|(label| |visualizer| |coordinate)|. Independently of the strategy, once |
| 1299 |
the data point has been chosen, the coordinate of the next data point is |
| 1300 |
stored in |(label| |visualizer| |coordinate')|. Then, a (conceptual) line |
| 1301 |
is created from the first coordinate to the second and a node is placed at |
| 1302 |
the beginning of this line to its ``left'' or, for the |text'| option, on |
| 1303 |
its ``right''. More precisely, an automatic anchor is computed for a node |
| 1304 |
placed implicitly on this line using the |auto| option or, for the |text'| |
| 1305 |
option, using |auto,swap|. |
| 1306 |
|
| 1307 |
The node placed at the position computed in this way will have the |
| 1308 |
\meta{text} set by the |text| or |text'| option and its styling is |
| 1309 |
determined by the current |node style|. |
| 1310 |
|
| 1311 |
Let us now have a look at the different ways of determining the data point |
| 1312 |
at which the label in anchored: |
| 1313 |
% |
| 1314 |
\begin{key}{/tikz/data visualization/visualizer label options/when=\meta{attribute}| is|\meta{number}} |
| 1315 |
This key causes the value of the \meta{attribute} to be monitored in |
| 1316 |
the stream of data points. The chosen is data point is the first data |
| 1317 |
point where the \meta{attribute} is at least \meta{number} (if this |
| 1318 |
never happens, the last data point is used). |
| 1319 |
% |
| 1320 |
\begin{codeexample}[ |
| 1321 |
width=6.3cm, |
| 1322 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 1323 |
pre={\tikz \datavisualization data group {function classes} = { |
| 1324 |
data [set=log, format=function] { |
| 1325 |
var x : interval [0.2:2.5]; |
| 1326 |
func y = ln(\value x); |
| 1327 |
} |
| 1328 |
data [set=lin, format=function] { |
| 1329 |
var x : interval [-2:2.5]; |
| 1330 |
func y = 0.5*\value x; |
| 1331 |
} |
| 1332 |
data [set=squared, format=function] { |
| 1333 |
var x : interval [-1.5:1.5]; |
| 1334 |
func y = \value x*\value x; |
| 1335 |
} |
| 1336 |
data [set=exp, format=function] { |
| 1337 |
var x : interval [-2.5:1]; |
| 1338 |
func y = exp(\value x); |
| 1339 |
} |
| 1340 |
};}, |
| 1341 |
] |
| 1342 |
\tikz \datavisualization [ |
| 1343 |
school book axes, |
| 1344 |
x axis={label=$x$}, |
| 1345 |
visualize as smooth line/.list={log, lin, squared, exp}, |
| 1346 |
log= {label in data={text'=$\log x$, when=y is -1, |
| 1347 |
text colored}}, |
| 1348 |
lin= {label in data={text=$x/2$, when=x is 2}}, |
| 1349 |
squared={label in data={text=$x^2$, when=x is 1.1}}, |
| 1350 |
exp= {label in data={text=$e^x$, when=x is -2, |
| 1351 |
text colored}}, |
| 1352 |
style sheet=vary hue] |
| 1353 |
data group {function classes}; |
| 1354 |
\end{codeexample} |
| 1355 |
\end{key} |
| 1356 |
% |
| 1357 |
\begin{key}{/tikz/data visualization/visualizer label options/index=\meta{number}} |
| 1358 |
This key chooses the \meta{number}th data point belonging to the |
| 1359 |
visualizer's data set. |
| 1360 |
% |
| 1361 |
\begin{codeexample}[ |
| 1362 |
width=6.3cm, |
| 1363 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 1364 |
pre={\tikz \datavisualization data group {function classes} = { |
| 1365 |
data [set=log, format=function] { |
| 1366 |
var x : interval [0.2:2.5]; |
| 1367 |
func y = ln(\value x); |
| 1368 |
} |
| 1369 |
data [set=lin, format=function] { |
| 1370 |
var x : interval [-2:2.5]; |
| 1371 |
func y = 0.5*\value x; |
| 1372 |
} |
| 1373 |
data [set=squared, format=function] { |
| 1374 |
var x : interval [-1.5:1.5]; |
| 1375 |
func y = \value x*\value x; |
| 1376 |
} |
| 1377 |
data [set=exp, format=function] { |
| 1378 |
var x : interval [-2.5:1]; |
| 1379 |
func y = exp(\value x); |
| 1380 |
} |
| 1381 |
};}, |
| 1382 |
] |
| 1383 |
\tikz \datavisualization [ |
| 1384 |
school book axes, |
| 1385 |
x axis={label=$x$}, |
| 1386 |
visualize as smooth line/.list={exp}, |
| 1387 |
exp= {label in data={text=$5$, index=5}, |
| 1388 |
label in data={text=$10$, index=10}, |
| 1389 |
label in data={text=$20$, index=20}, |
| 1390 |
style={mark=x}}, |
| 1391 |
style sheet=vary hue] |
| 1392 |
data group {function classes}; |
| 1393 |
\end{codeexample} |
| 1394 |
\end{key} |
| 1395 |
% |
| 1396 |
\begin{key}{/tikz/data visualization/visualizer label options/pos=\meta{fraction}} |
| 1397 |
This key chooses the first data point belonging to the data set whose |
| 1398 |
index is at least \meta{fraction} times the number of all data points |
| 1399 |
in the data set. |
| 1400 |
% |
| 1401 |
\begin{codeexample}[ |
| 1402 |
width=6.3cm, |
| 1403 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 1404 |
pre={\tikz \datavisualization data group {function classes} = { |
| 1405 |
data [set=log, format=function] { |
| 1406 |
var x : interval [0.2:2.5]; |
| 1407 |
func y = ln(\value x); |
| 1408 |
} |
| 1409 |
data [set=lin, format=function] { |
| 1410 |
var x : interval [-2:2.5]; |
| 1411 |
func y = 0.5*\value x; |
| 1412 |
} |
| 1413 |
data [set=squared, format=function] { |
| 1414 |
var x : interval [-1.5:1.5]; |
| 1415 |
func y = \value x*\value x; |
| 1416 |
} |
| 1417 |
data [set=exp, format=function] { |
| 1418 |
var x : interval [-2.5:1]; |
| 1419 |
func y = exp(\value x); |
| 1420 |
} |
| 1421 |
};}, |
| 1422 |
] |
| 1423 |
\tikz \datavisualization [ |
| 1424 |
school book axes, |
| 1425 |
x axis={label=$x$}, |
| 1426 |
visualize as smooth line=exp, |
| 1427 |
exp= {label in data={text=$.2$, pos=0.2}, |
| 1428 |
label in data={text=$.5$, pos=0.5}, |
| 1429 |
label in data={text=$.95$, pos=0.95}, |
| 1430 |
style={mark=x}}, |
| 1431 |
style sheet=vary hue] |
| 1432 |
data group {function classes}; |
| 1433 |
\end{codeexample} |
| 1434 |
\end{key} |
| 1435 |
% |
| 1436 |
\begin{key}{/tikz/data visualization/visualizer label options/auto} |
| 1437 |
This key is executed automatically by default. It works like the |pos| |
| 1438 |
option, where the \meta{fraction} is set to $(\meta{data set's |
| 1439 |
index}-1/2)/\meta{number of data sets}$. For instance, when there are |
| 1440 |
$10$ data sets, the fraction for the first one will be $5\%$, the |
| 1441 |
fraction for the second will be $15\%$, for the third it will be |
| 1442 |
$25\%$, ending with $95\%$ for the last one. |
| 1443 |
|
| 1444 |
The net effect of all this is that when there are several lines, labels |
| 1445 |
will be placed at different positions along the lines with hopefully |
| 1446 |
only little overlap. |
| 1447 |
% |
| 1448 |
\begin{codeexample}[ |
| 1449 |
width=6.3cm, |
| 1450 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 1451 |
] |
| 1452 |
\tikz \datavisualization [ |
| 1453 |
scientific axes=clean, |
| 1454 |
visualize as smooth line/.list={linear, squared, cubed}, |
| 1455 |
linear ={label in data={text=$2x$}}, |
| 1456 |
squared={label in data={text=$x^2$}}, |
| 1457 |
cubed ={label in data={text=$x^3$}}] |
| 1458 |
data [set=linear, format=function] { |
| 1459 |
var x : interval [0:1.5]; |
| 1460 |
func y = 2*\value x; |
| 1461 |
} |
| 1462 |
data [set=squared, format=function] { |
| 1463 |
var x : interval [0:1.5]; |
| 1464 |
func y = \value x * \value x; |
| 1465 |
} |
| 1466 |
data [set=cubed, format=function] { |
| 1467 |
var x : interval [0:1.5]; |
| 1468 |
func y = \value x * \value x * \value x; |
| 1469 |
}; |
| 1470 |
\end{codeexample} |
| 1471 |
% |
| 1472 |
As can be seen in the example, the result is not always satisfactory. |
| 1473 |
In this case, the |pin in data| option might be preferable, see below. |
| 1474 |
\end{key} |
| 1475 |
|
| 1476 |
The following keys allow you to style labels. |
| 1477 |
|
| 1478 |
\begin{key}{/tikz/data visualization/visualizer label options/node style=\meta{options}} |
| 1479 |
Just passes the options to |/tikz/data visualization/node style|. |
| 1480 |
\end{key} |
| 1481 |
% |
| 1482 |
\begin{key}{/tikz/data visualization/visualizer label options/text colored} |
| 1483 |
Causes the |node style| to set the text color to |visualizer color|. |
| 1484 |
The effect of this is that the label's text will have the same color as |
| 1485 |
the data set to which it is attached. |
| 1486 |
\end{key} |
| 1487 |
|
| 1488 |
\begin{stylekey}{/tikz/data visualization/every data set label} |
| 1489 |
This style is executed with every label that represents a data set. |
| 1490 |
Inside this style, use |node style| to change the appearance of nodes. |
| 1491 |
This style has a default definition, usually you should just append |
| 1492 |
things to this style. |
| 1493 |
% |
| 1494 |
\begin{codeexample}[ |
| 1495 |
width=6.3cm, |
| 1496 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 1497 |
pre={\tikz \datavisualization data group {function classes} = { |
| 1498 |
data [set=log, format=function] { |
| 1499 |
var x : interval [0.2:2.5]; |
| 1500 |
func y = ln(\value x); |
| 1501 |
} |
| 1502 |
data [set=lin, format=function] { |
| 1503 |
var x : interval [-2:2.5]; |
| 1504 |
func y = 0.5*\value x; |
| 1505 |
} |
| 1506 |
data [set=squared, format=function] { |
| 1507 |
var x : interval [-1.5:1.5]; |
| 1508 |
func y = \value x*\value x; |
| 1509 |
} |
| 1510 |
data [set=exp, format=function] { |
| 1511 |
var x : interval [-2.5:1]; |
| 1512 |
func y = exp(\value x); |
| 1513 |
} |
| 1514 |
};}, |
| 1515 |
] |
| 1516 |
\tikz \datavisualization [ |
| 1517 |
school book axes, |
| 1518 |
x axis={label=$x$}, |
| 1519 |
visualize as smooth line/.list={log, lin, squared, exp}, |
| 1520 |
every data set label/.append style={text colored}, |
| 1521 |
log= {label in data={text'=$\log x$, when=y is -1}}, |
| 1522 |
lin= {label in data={text=$x/2$, |
| 1523 |
node style=sloped, when=x is 2}}, |
| 1524 |
squared={label in data={text=$x^2$, when=x is 1.1}}, |
| 1525 |
exp= {label in data={text=$e^x$, |
| 1526 |
node style=sloped, when=x is -2}}, |
| 1527 |
style sheet=vary hue] |
| 1528 |
data group {function classes}; |
| 1529 |
\end{codeexample} |
| 1530 |
\end{stylekey} |
| 1531 |
|
| 1532 |
\begin{stylekey}{/tikz/data visualization/every label in data} |
| 1533 |
Like |every data set label|, this key is also executed with labels. |
| 1534 |
However, this key is executed after the style sheets have been |
| 1535 |
executed, giving you a chance to overrule their styling. |
| 1536 |
\end{stylekey} |
| 1537 |
\end{key} |
| 1538 |
|
| 1539 |
|
| 1540 |
\subsubsection{Connecting a Label to a Data Set via a Pin} |
| 1541 |
|
| 1542 |
\begin{key}{/tikz/data visualization/visualizer options/pin in data=\meta{options}} |
| 1543 |
This key is a variant of the |label in data| key and takes the same |
| 1544 |
options, plus two additional ones. The difference to |label in data| is |
| 1545 |
that the label node is shown a bit removed from the data set, but connected |
| 1546 |
to it via a small line (this is like the difference between the |label| and |
| 1547 |
|pin| options). |
| 1548 |
% |
| 1549 |
\begin{codeexample}[ |
| 1550 |
width=6.3cm, |
| 1551 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 1552 |
] |
| 1553 |
\tikz \datavisualization [ |
| 1554 |
scientific axes=clean, |
| 1555 |
visualize as smooth line/.list={linear, squared, cubed}, |
| 1556 |
linear ={pin in data={text=$2x$}}, |
| 1557 |
squared={pin in data={text=$x^2$}}, |
| 1558 |
cubed ={pin in data={text=$x^3$}}] |
| 1559 |
data [set=linear, format=function] { |
| 1560 |
var x : interval [0:1.5]; |
| 1561 |
func y = \value x; |
| 1562 |
} |
| 1563 |
data [set=squared, format=function] { |
| 1564 |
var x : interval [0:1.5]; |
| 1565 |
func y = \value x * \value x; |
| 1566 |
} |
| 1567 |
data [set=cubed, format=function] { |
| 1568 |
var x : interval [0:1.5]; |
| 1569 |
func y = \value x * \value x * \value x; |
| 1570 |
}; |
| 1571 |
\end{codeexample} |
| 1572 |
% |
| 1573 |
The following keys can be used additionally: |
| 1574 |
% |
| 1575 |
\begin{key}{/tikz/data visualization/visualizer label options/pin angle=\meta{angle}} |
| 1576 |
The position of the label of a |pin in data| is mainly computed in the |
| 1577 |
same way as for a |label in data|. However, once the position has been |
| 1578 |
computed, the label is shifted as follows: |
| 1579 |
% |
| 1580 |
\begin{itemize} |
| 1581 |
\item When an \meta{angle} is specified using the present key, the |
| 1582 |
shift is by the current value of |pin length| in the direction |
| 1583 |
of \meta{angle}. |
| 1584 |
\item When \meta{angle} is empty (which is the default), then the |
| 1585 |
shift is also by the current value of |pin length|, but now in |
| 1586 |
the direction that is orthogonal and to the left of the line |
| 1587 |
between the coordinate of the data point and the coordinate of |
| 1588 |
the next data point. When |text'| is used, the direction is to |
| 1589 |
the right instead of the left. |
| 1590 |
\end{itemize} |
| 1591 |
\end{key} |
| 1592 |
|
| 1593 |
\begin{key}{/tikz/data visualization/visualizer label options/pin length=\meta{dimension}} |
| 1594 |
See the description of |pin angle|. |
| 1595 |
\end{key} |
| 1596 |
% |
| 1597 |
\begin{codeexample}[ |
| 1598 |
width=6.3cm, |
| 1599 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 1600 |
pre={\tikz \datavisualization data group {function classes} = { |
| 1601 |
data [set=log, format=function] { |
| 1602 |
var x : interval [0.2:2.5]; |
| 1603 |
func y = ln(\value x); |
| 1604 |
} |
| 1605 |
data [set=lin, format=function] { |
| 1606 |
var x : interval [-2:2.5]; |
| 1607 |
func y = 0.5*\value x; |
| 1608 |
} |
| 1609 |
data [set=squared, format=function] { |
| 1610 |
var x : interval [-1.5:1.5]; |
| 1611 |
func y = \value x*\value x; |
| 1612 |
} |
| 1613 |
data [set=exp, format=function] { |
| 1614 |
var x : interval [-2.5:1]; |
| 1615 |
func y = exp(\value x); |
| 1616 |
} |
| 1617 |
};}, |
| 1618 |
] |
| 1619 |
\tikz \datavisualization [ |
| 1620 |
school book axes, |
| 1621 |
x axis={label=$x$}, |
| 1622 |
visualize as smooth line/.list={log, lin, squared, exp}, |
| 1623 |
every data set label/.append style={text colored}, |
| 1624 |
log= {pin in data={text'=$\log x$, when=y is -1}}, |
| 1625 |
lin= {pin in data={text=$x/2$, when=x is 2, |
| 1626 |
pin length=1ex}}, |
| 1627 |
squared={pin in data={text=$x^2$, when=x is 1.1, |
| 1628 |
pin angle=230}}, |
| 1629 |
exp= {label in data={text=$e^x$, when=x is -2}}, |
| 1630 |
style sheet=vary hue] |
| 1631 |
data group {function classes}; |
| 1632 |
\end{codeexample} |
| 1633 |
\end{key} |
| 1634 |
|
| 1635 |
|
| 1636 |
\subsection{Usage: Labeling Data Sets Inside a Legend} |
| 1637 |
|
| 1638 |
The ``classical'' way of indicating the style used for the different data sets |
| 1639 |
inside a visualization is a \emph{legend}. It is a description next to or even |
| 1640 |
inside the visualization that contains one line for each data set and displays |
| 1641 |
an iconographic version of the data set next to some text labeling the data |
| 1642 |
set. Note, however, that even though legend are quite common, also consider |
| 1643 |
using a |label in data| or a |pin in data| instead. |
| 1644 |
|
| 1645 |
Creating a high-quality legend is by no means simple. A legend should not |
| 1646 |
distract the reader, so aggressive borders should definitively be avoided. A |
| 1647 |
legend should make it easy to match the actual styling of a data set (like, |
| 1648 |
say, using a red, dashed line) to the ``iconographic'' representation of this |
| 1649 |
styling. An example of what can go wrong here is using short lines to represent |
| 1650 |
lines dashed in different way where the lines are so short that the differences |
| 1651 |
in the dashing cannot be discerned. Another example is showing straight lines |
| 1652 |
with plot marks on them where the plot marks are obscured by the horizontal |
| 1653 |
line itself, while the plot marks are clearly visible in the actual |
| 1654 |
visualization since no horizontal lines occur. |
| 1655 |
|
| 1656 |
The data visualization engine comes with a large set of options for creating |
| 1657 |
and placing high-quality legends next or inside data visualizations. |
| 1658 |
|
| 1659 |
|
| 1660 |
\subsubsection{Creating Legends and Legend Entries} |
| 1661 |
|
| 1662 |
A data visualization can be accompanied by one or more legends. In order to |
| 1663 |
create a legend, the following key can be used (although, in practice, you will |
| 1664 |
usually use the |legend| key instead, see below): |
| 1665 |
|
| 1666 |
\begin{key}{/tikz/data visualization/new legend=\meta{legend name} (default main legend)} |
| 1667 |
This key is used to create a new legend named \meta{legend name}. The |
| 1668 |
legend is empty by default and further options are needed to add entries to |
| 1669 |
it. When the key is called a second time for the same \meta{legend name} |
| 1670 |
nothing happens. |
| 1671 |
|
| 1672 |
When a legend is created, a new key is created that can subsequently be |
| 1673 |
used to configure the legend: |
| 1674 |
% |
| 1675 |
\begin{key}{/tikz/data visualization/\meta{legend name}=\meta{options}} |
| 1676 |
When this key is used, the \meta{options} are executed with the path |
| 1677 |
prefix |
| 1678 |
% |
| 1679 |
\begin{codeexample}[code only] |
| 1680 |
/tikz/data visualization/legend options |
| 1681 |
\end{codeexample} |
| 1682 |
% |
| 1683 |
The different keys with this path prefix allow you to change the |
| 1684 |
position where the legend is shown and how it is organised (for |
| 1685 |
instance, whether legend entries are shown in a row or in a column or |
| 1686 |
in a square). |
| 1687 |
|
| 1688 |
The different possible keys will be explained in the course of this |
| 1689 |
section. |
| 1690 |
\end{key} |
| 1691 |
|
| 1692 |
In the end, the legend is just a \tikzname\ node, a |matrix| node, to be |
| 1693 |
precise. The following key is used to style this node: |
| 1694 |
|
| 1695 |
\begin{key}{/tikz/data visualization/legend options/matrix node style=\meta{options}} |
| 1696 |
Adds the \meta{options} to the list of options that will be executed |
| 1697 |
when the legend's node is created. |
| 1698 |
% |
| 1699 |
\begin{codeexample}[ |
| 1700 |
width=8cm, |
| 1701 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 1702 |
pre={\tikz \datavisualization data group {function classes} = { |
| 1703 |
data [set=log, format=function] { |
| 1704 |
var x : interval [0.2:2.5]; |
| 1705 |
func y = ln(\value x); |
| 1706 |
} |
| 1707 |
data [set=lin, format=function] { |
| 1708 |
var x : interval [-2:2.5]; |
| 1709 |
func y = 0.5*\value x; |
| 1710 |
} |
| 1711 |
data [set=squared, format=function] { |
| 1712 |
var x : interval [-1.5:1.5]; |
| 1713 |
func y = \value x*\value x; |
| 1714 |
} |
| 1715 |
data [set=exp, format=function] { |
| 1716 |
var x : interval [-2.5:1]; |
| 1717 |
func y = exp(\value x); |
| 1718 |
} |
| 1719 |
};}, |
| 1720 |
] |
| 1721 |
\tikz \datavisualization [ |
| 1722 |
scientific axes, |
| 1723 |
visualize as smooth line/.list= |
| 1724 |
{log, lin, squared, exp}, |
| 1725 |
legend={matrix node style={fill=black!25}}, |
| 1726 |
log= {label in legend={text=$\log x$}}, |
| 1727 |
lin= {label in legend={text=$x/2$}}, |
| 1728 |
squared={label in legend={text=$x^2$}}, |
| 1729 |
exp= {label in legend={text=$e^x$}}, |
| 1730 |
style sheet=vary dashing] |
| 1731 |
data group {function classes}; |
| 1732 |
\end{codeexample} |
| 1733 |
\end{key} |
| 1734 |
|
| 1735 |
The following style allows you to configure the default appearance of every |
| 1736 |
newly created legend: |
| 1737 |
% |
| 1738 |
\begin{stylekey}{/tikz/data visualization/legend options/every new legend} |
| 1739 |
This key defaults to |east outside, label style=text right|. This means |
| 1740 |
that by default a legend is placed to the right of the data |
| 1741 |
visualization and that in the individual legend entries the text is to |
| 1742 |
the right of the data set visualization. |
| 1743 |
\end{stylekey} |
| 1744 |
% |
| 1745 |
\begin{codeexample}[ |
| 1746 |
width=6cm, |
| 1747 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 1748 |
pre={\tikz \datavisualization data group {function classes} = { |
| 1749 |
data [set=log, format=function] { |
| 1750 |
var x : interval [0.2:2.5]; |
| 1751 |
func y = ln(\value x); |
| 1752 |
} |
| 1753 |
data [set=lin, format=function] { |
| 1754 |
var x : interval [-2:2.5]; |
| 1755 |
func y = 0.5*\value x; |
| 1756 |
} |
| 1757 |
data [set=squared, format=function] { |
| 1758 |
var x : interval [-1.5:1.5]; |
| 1759 |
func y = \value x*\value x; |
| 1760 |
} |
| 1761 |
data [set=exp, format=function] { |
| 1762 |
var x : interval [-2.5:1]; |
| 1763 |
func y = exp(\value x); |
| 1764 |
} |
| 1765 |
};}, |
| 1766 |
] |
| 1767 |
\tikz \datavisualization [ |
| 1768 |
scientific axes, x axis={label=$x$}, |
| 1769 |
visualize as smooth line/.list={log, lin, squared, exp}, |
| 1770 |
new legend={upper legend}, |
| 1771 |
new legend={lower legend}, |
| 1772 |
upper legend=above, |
| 1773 |
lower legend=below, |
| 1774 |
log= {label in legend={text=$\log x$, legend=upper legend}}, |
| 1775 |
lin= {label in legend={text=$x/2$, legend=upper legend}}, |
| 1776 |
squared={label in legend={text=$x^2$, legend=lower legend}}, |
| 1777 |
exp= {label in legend={text=$e^x$, legend=lower legend}}, |
| 1778 |
style sheet=vary dashing] |
| 1779 |
data group {function classes}; |
| 1780 |
\end{codeexample} |
| 1781 |
% |
| 1782 |
\end{key} |
| 1783 |
|
| 1784 |
\begin{key}{/tikz/data visualization/legend=\meta{options}} |
| 1785 |
This is a shorthand for |
| 1786 |
|new legend=main legend, main legend=|\meta{options}. In other words, this |
| 1787 |
key creates a new |main legend| and immediately passes the configuration |
| 1788 |
\meta{options} to this legend. |
| 1789 |
% |
| 1790 |
\begin{codeexample}[ |
| 1791 |
width=7cm, |
| 1792 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 1793 |
pre={\tikz \datavisualization data group {function classes} = { |
| 1794 |
data [set=log, format=function] { |
| 1795 |
var x : interval [0.2:2.5]; |
| 1796 |
func y = ln(\value x); |
| 1797 |
} |
| 1798 |
data [set=lin, format=function] { |
| 1799 |
var x : interval [-2:2.5]; |
| 1800 |
func y = 0.5*\value x; |
| 1801 |
} |
| 1802 |
data [set=squared, format=function] { |
| 1803 |
var x : interval [-1.5:1.5]; |
| 1804 |
func y = \value x*\value x; |
| 1805 |
} |
| 1806 |
data [set=exp, format=function] { |
| 1807 |
var x : interval [-2.5:1]; |
| 1808 |
func y = exp(\value x); |
| 1809 |
} |
| 1810 |
};}, |
| 1811 |
] |
| 1812 |
\tikz \datavisualization [ |
| 1813 |
scientific axes, x axis={label=$x$}, |
| 1814 |
visualize as smooth line/.list={log, lin, squared, exp}, |
| 1815 |
legend=below, |
| 1816 |
log= {label in legend={text=$\log x$}}, |
| 1817 |
lin= {label in legend={text=$x/2$}}, |
| 1818 |
squared={label in legend={text=$x^2$}}, |
| 1819 |
exp= {label in legend={text=$e^x$}}, |
| 1820 |
style sheet=vary dashing] |
| 1821 |
data group {function classes}; |
| 1822 |
\end{codeexample} |
| 1823 |
% |
| 1824 |
\end{key} |
| 1825 |
|
| 1826 |
As pointed out above, a legend is empty by default. In particular, the |
| 1827 |
different data sets are not automatically inserted into the legend. Instead, |
| 1828 |
the key |label in legend| must be used together with a data set: |
| 1829 |
|
| 1830 |
\begin{key}{/tikz/data visualization/visualizer options/label in legend=\meta{options}} |
| 1831 |
This key is passed to a data set, similar to options like |pin in data| or |
| 1832 |
|smooth line|. The \meta{options} are used to configure the following: |
| 1833 |
% |
| 1834 |
\begin{itemize} |
| 1835 |
\item The legend in which the data set should be visualized. |
| 1836 |
\item The text that is to be shown in the legend for the data set. |
| 1837 |
\item The appearance of the legend entries. |
| 1838 |
\end{itemize} |
| 1839 |
% |
| 1840 |
In detail, the \meta{options} are executed with the path prefix |
| 1841 |
% |
| 1842 |
\begin{codeexample}[code only] |
| 1843 |
/tikz/data visualization/legend entry options |
| 1844 |
\end{codeexample} |
| 1845 |
% |
| 1846 |
To configure in which legend the label should appear, use the |
| 1847 |
following key: |
| 1848 |
% |
| 1849 |
\begin{key}{/tikz/data visualization/legend entry options/legend=\meta{name} (initially main legend)} |
| 1850 |
Set this key to the name of a legend that has previously been created |
| 1851 |
using |new legend|. The label will then be shown in this legend. |
| 1852 |
|
| 1853 |
In most cases, there is only one legend (namely |main legend|) and |
| 1854 |
there is no need to set this key since it defaults to the main legend. |
| 1855 |
|
| 1856 |
Also note that the legend \meta{name} is automatically created if it |
| 1857 |
nodes not yet exist. |
| 1858 |
\end{key} |
| 1859 |
|
| 1860 |
\begin{key}{/tikz/data visualization/legend entry options/text=\meta{text}} |
| 1861 |
Use this key to setup the \meta{text} that is shown as the label of the |
| 1862 |
data set. |
| 1863 |
% |
| 1864 |
\begin{codeexample}[ |
| 1865 |
width=8cm, |
| 1866 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 1867 |
pre={\tikz \datavisualization data group {function classes} = { |
| 1868 |
data [set=log, format=function] { |
| 1869 |
var x : interval [0.2:2.5]; |
| 1870 |
func y = ln(\value x); |
| 1871 |
} |
| 1872 |
data [set=lin, format=function] { |
| 1873 |
var x : interval [-2:2.5]; |
| 1874 |
func y = 0.5*\value x; |
| 1875 |
} |
| 1876 |
data [set=squared, format=function] { |
| 1877 |
var x : interval [-1.5:1.5]; |
| 1878 |
func y = \value x*\value x; |
| 1879 |
} |
| 1880 |
data [set=exp, format=function] { |
| 1881 |
var x : interval [-2.5:1]; |
| 1882 |
func y = exp(\value x); |
| 1883 |
} |
| 1884 |
};}, |
| 1885 |
] |
| 1886 |
\tikz \datavisualization [ |
| 1887 |
scientific axes, x axis={label=$x$}, |
| 1888 |
visualize as smooth line/.list= |
| 1889 |
{log, lin, squared, exp}, |
| 1890 |
log= {label in legend={text=$\log x$}}, |
| 1891 |
lin= {label in legend={text=$x/2$}}, |
| 1892 |
squared={pin in data ={text=$x^2$, pos=0.1}}, |
| 1893 |
exp= {label in data ={text=$e^x$}}, |
| 1894 |
style sheet=vary dashing] |
| 1895 |
data group {function classes}; |
| 1896 |
\end{codeexample} |
| 1897 |
\end{key} |
| 1898 |
|
| 1899 |
In addition to the two keys described above, there are further keys that |
| 1900 |
are described in Section~\ref{section-dv-label-legend-entry-options}. |
| 1901 |
\end{key} |
| 1902 |
|
| 1903 |
|
| 1904 |
\subsubsection{Rows and Columns of Legend Entries} |
| 1905 |
|
| 1906 |
In a legend, the different legend entries are arranged in a matrix, which |
| 1907 |
typically has only one row or one column. For the impatient reader: Say |
| 1908 |
|rows=1| to get everything in a row, say |columns=1| to get everything in a |
| 1909 |
single column, and skip the rest of this section. |
| 1910 |
|
| 1911 |
The more patient reader will appreciate that when there are very many different |
| 1912 |
data sets in a single visualization, it may be necessary to use more than one |
| 1913 |
row or column inside the legend. \tikzname\ comes with a rather powerful |
| 1914 |
mechanism for distributing the multiple legend entries over the matrix. |
| 1915 |
|
| 1916 |
The first thing to decide is in which ``direction'' the entries should be |
| 1917 |
inserted into the matrix. Suppose we have a $3 \times 3$ matrix and our entries |
| 1918 |
are $a$, $b$, $c$, and so on. Then, one might place the $a$ in the upper left |
| 1919 |
corner of the matrix, $b$ in the upper middle position, $c$ in the upper right |
| 1920 |
position, and $d$ in the middle left position. This is a ``first right, then |
| 1921 |
down'' strategy. A different strategy might be to place the $a$ in the upper |
| 1922 |
left corner, but $b$ in the middle left position, $c$ in the lower left |
| 1923 |
position, and $d$ then in the upper middle position. This is a ``first down, |
| 1924 |
then right'' strategy. In certain situations it might even make sense to place |
| 1925 |
$a$ in the lower right corner and then go ``first up, then left''. |
| 1926 |
|
| 1927 |
All of these strategies are supported by the |legend| command. You can |
| 1928 |
configure which strategy is used using the following keys: |
| 1929 |
|
| 1930 |
\tikzdatavisualizationset { |
| 1931 |
legend example/.style={ |
| 1932 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 1933 |
1={label in legend={text=1}}, |
| 1934 |
2={label in legend={text=2}}, |
| 1935 |
3={label in legend={text=3}}, |
| 1936 |
4={label in legend={text=4}}, |
| 1937 |
5={label in legend={text=5}}, |
| 1938 |
6={label in legend={text=6}}, |
| 1939 |
7={label in legend={text=7}}, |
| 1940 |
8={label in legend={text=8}} |
| 1941 |
} |
| 1942 |
} |
| 1943 |
|
| 1944 |
\begin{key}{/tikz/data visualization/legend options/down then right} |
| 1945 |
Causes the legend entries to fill the legend matrix first downward and, |
| 1946 |
once a column is full, the next column is begun to the right of the |
| 1947 |
previous one. This is the default. |
| 1948 |
% |
| 1949 |
\begin{codeexample}[ |
| 1950 |
width=6cm, |
| 1951 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 1952 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 1953 |
data [format=function] { |
| 1954 |
var set : {1,...,8}; |
| 1955 |
var x : interval [0:50]; |
| 1956 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 1957 |
} |
| 1958 |
};% |
| 1959 |
\tikzdatavisualizationset { |
| 1960 |
legend example/.style={ |
| 1961 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 1962 |
1={label in legend={text=1}}, |
| 1963 |
2={label in legend={text=2}}, |
| 1964 |
3={label in legend={text=3}}, |
| 1965 |
4={label in legend={text=4}}, |
| 1966 |
5={label in legend={text=5}}, |
| 1967 |
6={label in legend={text=6}}, |
| 1968 |
7={label in legend={text=7}}, |
| 1969 |
8={label in legend={text=8}} |
| 1970 |
} |
| 1971 |
}}, |
| 1972 |
] |
| 1973 |
\tikz \datavisualization [ |
| 1974 |
visualize as smooth line/.list={1,2,3,4,5,6,7,8}, |
| 1975 |
legend example, style sheet=vary hue, |
| 1976 |
main legend={down then right, columns=3}] |
| 1977 |
data group {sin functions}; |
| 1978 |
\end{codeexample} |
| 1979 |
% |
| 1980 |
In the example, the |legend example| is the following style: |
| 1981 |
% |
| 1982 |
\begin{codeexample}[code only] |
| 1983 |
\tikzdatavisualizationset { |
| 1984 |
legend example/.style={ |
| 1985 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 1986 |
1={label in legend={text=1}}, |
| 1987 |
2={label in legend={text=2}}, |
| 1988 |
3={label in legend={text=3}}, |
| 1989 |
4={label in legend={text=4}}, |
| 1990 |
5={label in legend={text=5}}, |
| 1991 |
6={label in legend={text=6}}, |
| 1992 |
7={label in legend={text=7}}, |
| 1993 |
8={label in legend={text=8}} |
| 1994 |
} |
| 1995 |
} |
| 1996 |
\end{codeexample} |
| 1997 |
% |
| 1998 |
\end{key} |
| 1999 |
|
| 2000 |
\begin{key}{/tikz/data visualization/legend options/down then left} |
| 2001 |
\begin{codeexample}[ |
| 2002 |
width=6cm, |
| 2003 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2004 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 2005 |
data [format=function] { |
| 2006 |
var set : {1,...,8}; |
| 2007 |
var x : interval [0:50]; |
| 2008 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 2009 |
} |
| 2010 |
};% |
| 2011 |
\tikzdatavisualizationset { |
| 2012 |
legend example/.style={ |
| 2013 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 2014 |
1={label in legend={text=1}}, |
| 2015 |
2={label in legend={text=2}}, |
| 2016 |
3={label in legend={text=3}}, |
| 2017 |
4={label in legend={text=4}}, |
| 2018 |
5={label in legend={text=5}}, |
| 2019 |
6={label in legend={text=6}}, |
| 2020 |
7={label in legend={text=7}}, |
| 2021 |
8={label in legend={text=8}} |
| 2022 |
} |
| 2023 |
}}, |
| 2024 |
] |
| 2025 |
\tikz \datavisualization [ |
| 2026 |
visualize as smooth line/.list={1,2,3,4,5,6,7,8}, |
| 2027 |
legend example, style sheet=vary hue, |
| 2028 |
main legend={down then left, columns=3}] |
| 2029 |
data group {sin functions}; |
| 2030 |
\end{codeexample} |
| 2031 |
\end{key} |
| 2032 |
|
| 2033 |
\begin{key}{/tikz/data visualization/legend options/up then right} |
| 2034 |
\begin{codeexample}[ |
| 2035 |
width=6cm, |
| 2036 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2037 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 2038 |
data [format=function] { |
| 2039 |
var set : {1,...,8}; |
| 2040 |
var x : interval [0:50]; |
| 2041 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 2042 |
} |
| 2043 |
};% |
| 2044 |
\tikzdatavisualizationset { |
| 2045 |
legend example/.style={ |
| 2046 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 2047 |
1={label in legend={text=1}}, |
| 2048 |
2={label in legend={text=2}}, |
| 2049 |
3={label in legend={text=3}}, |
| 2050 |
4={label in legend={text=4}}, |
| 2051 |
5={label in legend={text=5}}, |
| 2052 |
6={label in legend={text=6}}, |
| 2053 |
7={label in legend={text=7}}, |
| 2054 |
8={label in legend={text=8}} |
| 2055 |
} |
| 2056 |
}}, |
| 2057 |
] |
| 2058 |
\tikz \datavisualization [ |
| 2059 |
visualize as smooth line/.list={1,2,3,4,5,6,7,8}, |
| 2060 |
legend example, style sheet=vary hue, |
| 2061 |
main legend={up then right, columns=3}] |
| 2062 |
data group {sin functions}; |
| 2063 |
\end{codeexample} |
| 2064 |
\end{key} |
| 2065 |
|
| 2066 |
\begin{key}{/tikz/data visualization/legend options/up then left} |
| 2067 |
\begin{codeexample}[ |
| 2068 |
width=6cm, |
| 2069 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2070 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 2071 |
data [format=function] { |
| 2072 |
var set : {1,...,8}; |
| 2073 |
var x : interval [0:50]; |
| 2074 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 2075 |
} |
| 2076 |
};% |
| 2077 |
\tikzdatavisualizationset { |
| 2078 |
legend example/.style={ |
| 2079 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 2080 |
1={label in legend={text=1}}, |
| 2081 |
2={label in legend={text=2}}, |
| 2082 |
3={label in legend={text=3}}, |
| 2083 |
4={label in legend={text=4}}, |
| 2084 |
5={label in legend={text=5}}, |
| 2085 |
6={label in legend={text=6}}, |
| 2086 |
7={label in legend={text=7}}, |
| 2087 |
8={label in legend={text=8}} |
| 2088 |
} |
| 2089 |
}}, |
| 2090 |
] |
| 2091 |
\tikz \datavisualization [ |
| 2092 |
visualize as smooth line/.list={1,2,3,4,5,6,7,8}, |
| 2093 |
legend example, style sheet=vary hue, |
| 2094 |
main legend={up then left, columns=3}] |
| 2095 |
data group {sin functions}; |
| 2096 |
\end{codeexample} |
| 2097 |
\end{key} |
| 2098 |
|
| 2099 |
\begin{key}{/tikz/data visualization/legend options/left then up} |
| 2100 |
\begin{codeexample}[ |
| 2101 |
width=6cm, |
| 2102 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2103 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 2104 |
data [format=function] { |
| 2105 |
var set : {1,...,8}; |
| 2106 |
var x : interval [0:50]; |
| 2107 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 2108 |
} |
| 2109 |
};% |
| 2110 |
\tikzdatavisualizationset { |
| 2111 |
legend example/.style={ |
| 2112 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 2113 |
1={label in legend={text=1}}, |
| 2114 |
2={label in legend={text=2}}, |
| 2115 |
3={label in legend={text=3}}, |
| 2116 |
4={label in legend={text=4}}, |
| 2117 |
5={label in legend={text=5}}, |
| 2118 |
6={label in legend={text=6}}, |
| 2119 |
7={label in legend={text=7}}, |
| 2120 |
8={label in legend={text=8}} |
| 2121 |
} |
| 2122 |
}}, |
| 2123 |
] |
| 2124 |
\tikz \datavisualization [ |
| 2125 |
visualize as smooth line/.list={1,2,3,4,5,6,7,8}, |
| 2126 |
legend example, style sheet=vary hue, |
| 2127 |
main legend={left then up, columns=3}] |
| 2128 |
data group {sin functions}; |
| 2129 |
\end{codeexample} |
| 2130 |
\end{key} |
| 2131 |
|
| 2132 |
\begin{key}{/tikz/data visualization/legend options/left then down} |
| 2133 |
\begin{codeexample}[ |
| 2134 |
width=6cm, |
| 2135 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2136 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 2137 |
data [format=function] { |
| 2138 |
var set : {1,...,8}; |
| 2139 |
var x : interval [0:50]; |
| 2140 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 2141 |
} |
| 2142 |
};% |
| 2143 |
\tikzdatavisualizationset { |
| 2144 |
legend example/.style={ |
| 2145 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 2146 |
1={label in legend={text=1}}, |
| 2147 |
2={label in legend={text=2}}, |
| 2148 |
3={label in legend={text=3}}, |
| 2149 |
4={label in legend={text=4}}, |
| 2150 |
5={label in legend={text=5}}, |
| 2151 |
6={label in legend={text=6}}, |
| 2152 |
7={label in legend={text=7}}, |
| 2153 |
8={label in legend={text=8}} |
| 2154 |
} |
| 2155 |
}}, |
| 2156 |
] |
| 2157 |
\tikz \datavisualization [ |
| 2158 |
visualize as smooth line/.list={1,2,3,4,5,6,7,8}, |
| 2159 |
legend example, style sheet=vary hue, |
| 2160 |
main legend={left then down, columns=3}] |
| 2161 |
data group {sin functions}; |
| 2162 |
\end{codeexample} |
| 2163 |
\end{key} |
| 2164 |
|
| 2165 |
\begin{key}{/tikz/data visualization/legend options/right then up} |
| 2166 |
\begin{codeexample}[ |
| 2167 |
width=6cm, |
| 2168 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2169 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 2170 |
data [format=function] { |
| 2171 |
var set : {1,...,8}; |
| 2172 |
var x : interval [0:50]; |
| 2173 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 2174 |
} |
| 2175 |
};% |
| 2176 |
\tikzdatavisualizationset { |
| 2177 |
legend example/.style={ |
| 2178 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 2179 |
1={label in legend={text=1}}, |
| 2180 |
2={label in legend={text=2}}, |
| 2181 |
3={label in legend={text=3}}, |
| 2182 |
4={label in legend={text=4}}, |
| 2183 |
5={label in legend={text=5}}, |
| 2184 |
6={label in legend={text=6}}, |
| 2185 |
7={label in legend={text=7}}, |
| 2186 |
8={label in legend={text=8}} |
| 2187 |
} |
| 2188 |
}}, |
| 2189 |
] |
| 2190 |
\tikz \datavisualization [ |
| 2191 |
visualize as smooth line/.list={1,2,3,4,5,6,7,8}, |
| 2192 |
legend example, style sheet=vary hue, |
| 2193 |
main legend={right then up, columns=3}] |
| 2194 |
data group {sin functions}; |
| 2195 |
\end{codeexample} |
| 2196 |
\end{key} |
| 2197 |
|
| 2198 |
\begin{key}{/tikz/data visualization/legend options/right then down} |
| 2199 |
\begin{codeexample}[ |
| 2200 |
width=6cm, |
| 2201 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2202 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 2203 |
data [format=function] { |
| 2204 |
var set : {1,...,8}; |
| 2205 |
var x : interval [0:50]; |
| 2206 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 2207 |
} |
| 2208 |
};% |
| 2209 |
\tikzdatavisualizationset { |
| 2210 |
legend example/.style={ |
| 2211 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 2212 |
1={label in legend={text=1}}, |
| 2213 |
2={label in legend={text=2}}, |
| 2214 |
3={label in legend={text=3}}, |
| 2215 |
4={label in legend={text=4}}, |
| 2216 |
5={label in legend={text=5}}, |
| 2217 |
6={label in legend={text=6}}, |
| 2218 |
7={label in legend={text=7}}, |
| 2219 |
8={label in legend={text=8}} |
| 2220 |
} |
| 2221 |
}}, |
| 2222 |
] |
| 2223 |
\tikz \datavisualization [ |
| 2224 |
visualize as smooth line/.list={1,2,3,4,5,6,7,8}, |
| 2225 |
legend example, style sheet=vary hue, |
| 2226 |
main legend={right then down, columns=3}] |
| 2227 |
data group {sin functions}; |
| 2228 |
\end{codeexample} |
| 2229 |
\end{key} |
| 2230 |
|
| 2231 |
Having configured the directions in which the matrix is being filled, you must |
| 2232 |
next setup the number of rows or columns that are to be shown. There are |
| 2233 |
actually two different ways of doing so. The first way is to specify a maximum |
| 2234 |
number of rows or columns. For instance, you might specify that there should be |
| 2235 |
at most ten rows to a column and when there are more, a new column should be |
| 2236 |
begun. This is achieved using the following keys: |
| 2237 |
|
| 2238 |
\begin{key}{/tikz/data visualization/legend options/max rows=\meta{number}} |
| 2239 |
As the legend matrix is being filled, whenever the number of rows in the |
| 2240 |
current column would exceed \meta{number}, a new column is started. |
| 2241 |
% |
| 2242 |
\begin{codeexample}[ |
| 2243 |
width=7cm, |
| 2244 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2245 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 2246 |
data [format=function] { |
| 2247 |
var set : {1,...,8}; |
| 2248 |
var x : interval [0:50]; |
| 2249 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 2250 |
} |
| 2251 |
};% |
| 2252 |
\tikzdatavisualizationset { |
| 2253 |
legend example/.style={ |
| 2254 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 2255 |
1={label in legend={text=1}}, |
| 2256 |
2={label in legend={text=2}}, |
| 2257 |
3={label in legend={text=3}}, |
| 2258 |
4={label in legend={text=4}}, |
| 2259 |
5={label in legend={text=5}}, |
| 2260 |
6={label in legend={text=6}}, |
| 2261 |
7={label in legend={text=7}}, |
| 2262 |
8={label in legend={text=8}} |
| 2263 |
} |
| 2264 |
}}, |
| 2265 |
] |
| 2266 |
\tikz \datavisualization [ |
| 2267 |
visualize as smooth line/.list={1,2,3,4,5,6,7,8}, |
| 2268 |
legend example, style sheet=vary hue, |
| 2269 |
main legend={max rows=3}] |
| 2270 |
data group {sin functions}; |
| 2271 |
\end{codeexample} |
| 2272 |
% |
| 2273 |
\begin{codeexample}[ |
| 2274 |
width=7cm, |
| 2275 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2276 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 2277 |
data [format=function] { |
| 2278 |
var set : {1,...,8}; |
| 2279 |
var x : interval [0:50]; |
| 2280 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 2281 |
} |
| 2282 |
};% |
| 2283 |
\tikzdatavisualizationset { |
| 2284 |
legend example/.style={ |
| 2285 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 2286 |
1={label in legend={text=1}}, |
| 2287 |
2={label in legend={text=2}}, |
| 2288 |
3={label in legend={text=3}}, |
| 2289 |
4={label in legend={text=4}}, |
| 2290 |
5={label in legend={text=5}}, |
| 2291 |
6={label in legend={text=6}}, |
| 2292 |
7={label in legend={text=7}}, |
| 2293 |
8={label in legend={text=8}} |
| 2294 |
} |
| 2295 |
}}, |
| 2296 |
] |
| 2297 |
\tikz \datavisualization [ |
| 2298 |
visualize as smooth line/.list={1,2,3,4,5,6,7,8}, |
| 2299 |
legend example, style sheet=vary hue, |
| 2300 |
main legend={max rows=4}] |
| 2301 |
data group {sin functions}; |
| 2302 |
\end{codeexample} |
| 2303 |
% |
| 2304 |
\begin{codeexample}[ |
| 2305 |
width=7cm, |
| 2306 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2307 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 2308 |
data [format=function] { |
| 2309 |
var set : {1,...,8}; |
| 2310 |
var x : interval [0:50]; |
| 2311 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 2312 |
} |
| 2313 |
};% |
| 2314 |
\tikzdatavisualizationset { |
| 2315 |
legend example/.style={ |
| 2316 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 2317 |
1={label in legend={text=1}}, |
| 2318 |
2={label in legend={text=2}}, |
| 2319 |
3={label in legend={text=3}}, |
| 2320 |
4={label in legend={text=4}}, |
| 2321 |
5={label in legend={text=5}}, |
| 2322 |
6={label in legend={text=6}}, |
| 2323 |
7={label in legend={text=7}}, |
| 2324 |
8={label in legend={text=8}} |
| 2325 |
} |
| 2326 |
}}, |
| 2327 |
] |
| 2328 |
\tikz \datavisualization [ |
| 2329 |
visualize as smooth line/.list={1,2,3,4,5,6,7,8}, |
| 2330 |
legend example, style sheet=vary hue, |
| 2331 |
main legend={max rows=5}] |
| 2332 |
data group {sin functions}; |
| 2333 |
\end{codeexample} |
| 2334 |
% |
| 2335 |
\end{key} |
| 2336 |
|
| 2337 |
\begin{key}{/tikz/data visualization/legend options/max columns=\meta{number}} |
| 2338 |
This key works like |max rows|, only now the number of columns is |
| 2339 |
monitored. Note that this strategy only really makes sense when the when |
| 2340 |
you use this key with a strategy that first goes left or right and then up |
| 2341 |
or down. |
| 2342 |
% |
| 2343 |
\begin{codeexample}[ |
| 2344 |
width=7cm, |
| 2345 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2346 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 2347 |
data [format=function] { |
| 2348 |
var set : {1,...,8}; |
| 2349 |
var x : interval [0:50]; |
| 2350 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 2351 |
} |
| 2352 |
};% |
| 2353 |
\tikzdatavisualizationset { |
| 2354 |
legend example/.style={ |
| 2355 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 2356 |
1={label in legend={text=1}}, |
| 2357 |
2={label in legend={text=2}}, |
| 2358 |
3={label in legend={text=3}}, |
| 2359 |
4={label in legend={text=4}}, |
| 2360 |
5={label in legend={text=5}}, |
| 2361 |
6={label in legend={text=6}}, |
| 2362 |
7={label in legend={text=7}}, |
| 2363 |
8={label in legend={text=8}} |
| 2364 |
} |
| 2365 |
}}, |
| 2366 |
] |
| 2367 |
\tikz \datavisualization [ |
| 2368 |
visualize as smooth line/.list={1,2,3,4,5,6,7,8}, |
| 2369 |
legend example, style sheet=vary hue, |
| 2370 |
main legend={right then down, max columns=2}] |
| 2371 |
data group {sin functions}; |
| 2372 |
\end{codeexample} |
| 2373 |
% |
| 2374 |
\begin{codeexample}[ |
| 2375 |
width=7cm, |
| 2376 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2377 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 2378 |
data [format=function] { |
| 2379 |
var set : {1,...,8}; |
| 2380 |
var x : interval [0:50]; |
| 2381 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 2382 |
} |
| 2383 |
};% |
| 2384 |
\tikzdatavisualizationset { |
| 2385 |
legend example/.style={ |
| 2386 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 2387 |
1={label in legend={text=1}}, |
| 2388 |
2={label in legend={text=2}}, |
| 2389 |
3={label in legend={text=3}}, |
| 2390 |
4={label in legend={text=4}}, |
| 2391 |
5={label in legend={text=5}}, |
| 2392 |
6={label in legend={text=6}}, |
| 2393 |
7={label in legend={text=7}}, |
| 2394 |
8={label in legend={text=8}} |
| 2395 |
} |
| 2396 |
}}, |
| 2397 |
] |
| 2398 |
\tikz \datavisualization [ |
| 2399 |
visualize as smooth line/.list={1,2,3,4,5,6,7,8}, |
| 2400 |
legend example, style sheet=vary hue, |
| 2401 |
main legend={right then down,max columns=3}] |
| 2402 |
data group {sin functions}; |
| 2403 |
\end{codeexample} |
| 2404 |
% |
| 2405 |
\begin{codeexample}[ |
| 2406 |
width=7cm, |
| 2407 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2408 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 2409 |
data [format=function] { |
| 2410 |
var set : {1,...,8}; |
| 2411 |
var x : interval [0:50]; |
| 2412 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 2413 |
} |
| 2414 |
};% |
| 2415 |
\tikzdatavisualizationset { |
| 2416 |
legend example/.style={ |
| 2417 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 2418 |
1={label in legend={text=1}}, |
| 2419 |
2={label in legend={text=2}}, |
| 2420 |
3={label in legend={text=3}}, |
| 2421 |
4={label in legend={text=4}}, |
| 2422 |
5={label in legend={text=5}}, |
| 2423 |
6={label in legend={text=6}}, |
| 2424 |
7={label in legend={text=7}}, |
| 2425 |
8={label in legend={text=8}} |
| 2426 |
} |
| 2427 |
}}, |
| 2428 |
] |
| 2429 |
\tikz \datavisualization [ |
| 2430 |
visualize as smooth line/.list={1,2,3,4,5,6,7,8}, |
| 2431 |
legend example, style sheet=vary hue, |
| 2432 |
main legend={right then down,max columns=4}] |
| 2433 |
data group {sin functions}; |
| 2434 |
\end{codeexample} |
| 2435 |
% |
| 2436 |
\end{key} |
| 2437 |
|
| 2438 |
The second way of specifying the number of entries in a row or column is to |
| 2439 |
specify an ``ideal number of rows or columns''. The idea is as follows: Suppose |
| 2440 |
that we use the standard strategy and would like to have everything in two |
| 2441 |
columns. Then if there are eight entries, the first four should go to the first |
| 2442 |
column, while the next four should go to the second column. If we have 20 |
| 2443 |
entries, the first ten should go the first column and the next ten to the |
| 2444 |
second, and so on. So, in general, the objective is to distribute the entries |
| 2445 |
evenly so the this ``ideal number of columns'' is reached. Only when there are |
| 2446 |
too few entries to achieve this or when the number of entries per column would |
| 2447 |
exceed the |max rows| value, will the number of columns deviate from this ideal |
| 2448 |
value. |
| 2449 |
|
| 2450 |
\begin{key}{/tikz/data visualization/legend options/ideal number of columns=\meta{number}} |
| 2451 |
Specifies, that the entries should be split into \meta{number} different |
| 2452 |
columns, whenever possible. However, when there would be more than the |
| 2453 |
|max rows| value of rows per column, more columns than the ideal number are |
| 2454 |
created. |
| 2455 |
% |
| 2456 |
\begin{codeexample}[ |
| 2457 |
width=7cm, |
| 2458 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2459 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 2460 |
data [format=function] { |
| 2461 |
var set : {1,...,8}; |
| 2462 |
var x : interval [0:50]; |
| 2463 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 2464 |
} |
| 2465 |
};% |
| 2466 |
\tikzdatavisualizationset { |
| 2467 |
legend example/.style={ |
| 2468 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 2469 |
1={label in legend={text=1}}, |
| 2470 |
2={label in legend={text=2}}, |
| 2471 |
3={label in legend={text=3}}, |
| 2472 |
4={label in legend={text=4}}, |
| 2473 |
5={label in legend={text=5}}, |
| 2474 |
6={label in legend={text=6}}, |
| 2475 |
7={label in legend={text=7}}, |
| 2476 |
8={label in legend={text=8}} |
| 2477 |
} |
| 2478 |
}}, |
| 2479 |
] |
| 2480 |
\tikz \datavisualization [ |
| 2481 |
visualize as smooth line/.list={1,2,3,4,5,6,7,8}, |
| 2482 |
legend example, style sheet=vary hue, |
| 2483 |
main legend={ideal number of columns=2}] |
| 2484 |
data group {sin functions}; |
| 2485 |
\end{codeexample} |
| 2486 |
% |
| 2487 |
\begin{codeexample}[ |
| 2488 |
width=7cm, |
| 2489 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2490 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 2491 |
data [format=function] { |
| 2492 |
var set : {1,...,8}; |
| 2493 |
var x : interval [0:50]; |
| 2494 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 2495 |
} |
| 2496 |
};% |
| 2497 |
\tikzdatavisualizationset { |
| 2498 |
legend example/.style={ |
| 2499 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 2500 |
1={label in legend={text=1}}, |
| 2501 |
2={label in legend={text=2}}, |
| 2502 |
3={label in legend={text=3}}, |
| 2503 |
4={label in legend={text=4}}, |
| 2504 |
5={label in legend={text=5}}, |
| 2505 |
6={label in legend={text=6}}, |
| 2506 |
7={label in legend={text=7}}, |
| 2507 |
8={label in legend={text=8}} |
| 2508 |
} |
| 2509 |
}}, |
| 2510 |
] |
| 2511 |
\tikz \datavisualization [ |
| 2512 |
visualize as smooth line/.list={1,2,3,4,5,6,7,8}, |
| 2513 |
legend example, style sheet=vary hue, |
| 2514 |
main legend={ideal number of columns=4}] |
| 2515 |
data group {sin functions}; |
| 2516 |
\end{codeexample} |
| 2517 |
% |
| 2518 |
\begin{codeexample}[ |
| 2519 |
width=7cm, |
| 2520 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2521 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 2522 |
data [format=function] { |
| 2523 |
var set : {1,...,8}; |
| 2524 |
var x : interval [0:50]; |
| 2525 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 2526 |
} |
| 2527 |
};% |
| 2528 |
\tikzdatavisualizationset { |
| 2529 |
legend example/.style={ |
| 2530 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 2531 |
1={label in legend={text=1}}, |
| 2532 |
2={label in legend={text=2}}, |
| 2533 |
3={label in legend={text=3}}, |
| 2534 |
4={label in legend={text=4}}, |
| 2535 |
5={label in legend={text=5}}, |
| 2536 |
6={label in legend={text=6}}, |
| 2537 |
7={label in legend={text=7}}, |
| 2538 |
8={label in legend={text=8}} |
| 2539 |
} |
| 2540 |
}}, |
| 2541 |
] |
| 2542 |
\tikz \datavisualization [ |
| 2543 |
visualize as smooth line/.list={1,2,3,4,5,6,7,8}, |
| 2544 |
legend example, style sheet=vary hue, |
| 2545 |
main legend={max rows=3,ideal number of columns=2}] |
| 2546 |
data group {sin functions}; |
| 2547 |
\end{codeexample} |
| 2548 |
% |
| 2549 |
\end{key} |
| 2550 |
|
| 2551 |
\begin{key}{/tikz/data visualization/legend options/rows=\meta{number}} |
| 2552 |
Shorthand for |ideal number of rows=|\meta{number}. |
| 2553 |
\end{key} |
| 2554 |
|
| 2555 |
\begin{key}{/tikz/data visualization/legend options/ideal number of rows=\meta{number}} |
| 2556 |
Works like |ideal number of columns|. |
| 2557 |
% |
| 2558 |
\begin{codeexample}[ |
| 2559 |
width=7cm, |
| 2560 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2561 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 2562 |
data [format=function] { |
| 2563 |
var set : {1,...,8}; |
| 2564 |
var x : interval [0:50]; |
| 2565 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 2566 |
} |
| 2567 |
};% |
| 2568 |
\tikzdatavisualizationset { |
| 2569 |
legend example/.style={ |
| 2570 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 2571 |
1={label in legend={text=1}}, |
| 2572 |
2={label in legend={text=2}}, |
| 2573 |
3={label in legend={text=3}}, |
| 2574 |
4={label in legend={text=4}}, |
| 2575 |
5={label in legend={text=5}}, |
| 2576 |
6={label in legend={text=6}}, |
| 2577 |
7={label in legend={text=7}}, |
| 2578 |
8={label in legend={text=8}} |
| 2579 |
} |
| 2580 |
}}, |
| 2581 |
] |
| 2582 |
\tikz \datavisualization [ |
| 2583 |
visualize as smooth line/.list={1,2,3,4,5,6,7,8}, |
| 2584 |
legend example, style sheet=vary hue, |
| 2585 |
main legend={ideal number of rows=2}] |
| 2586 |
data group {sin functions}; |
| 2587 |
\end{codeexample} |
| 2588 |
% |
| 2589 |
\begin{codeexample}[ |
| 2590 |
width=7cm, |
| 2591 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2592 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 2593 |
data [format=function] { |
| 2594 |
var set : {1,...,8}; |
| 2595 |
var x : interval [0:50]; |
| 2596 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 2597 |
} |
| 2598 |
};% |
| 2599 |
\tikzdatavisualizationset { |
| 2600 |
legend example/.style={ |
| 2601 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 2602 |
1={label in legend={text=1}}, |
| 2603 |
2={label in legend={text=2}}, |
| 2604 |
3={label in legend={text=3}}, |
| 2605 |
4={label in legend={text=4}}, |
| 2606 |
5={label in legend={text=5}}, |
| 2607 |
6={label in legend={text=6}}, |
| 2608 |
7={label in legend={text=7}}, |
| 2609 |
8={label in legend={text=8}} |
| 2610 |
} |
| 2611 |
}}, |
| 2612 |
] |
| 2613 |
\tikz \datavisualization [ |
| 2614 |
visualize as smooth line/.list={1,2,3,4,5,6,7,8}, |
| 2615 |
legend example, style sheet=vary hue, |
| 2616 |
main legend={ideal number of rows=4}] |
| 2617 |
data group {sin functions}; |
| 2618 |
\end{codeexample} |
| 2619 |
% |
| 2620 |
\begin{codeexample}[ |
| 2621 |
width=7cm, |
| 2622 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2623 |
pre={\tikz \datavisualization data group {sin functions} = { |
| 2624 |
data [format=function] { |
| 2625 |
var set : {1,...,8}; |
| 2626 |
var x : interval [0:50]; |
| 2627 |
func y = sin(\value x * (\value{set}+10))/(\value{set}+5); |
| 2628 |
} |
| 2629 |
};% |
| 2630 |
\tikzdatavisualizationset { |
| 2631 |
legend example/.style={ |
| 2632 |
scientific axes, all axes={length=1cm, ticks=none}, |
| 2633 |
1={label in legend={text=1}}, |
| 2634 |
2={label in legend={text=2}}, |
| 2635 |
3={label in legend={text=3}}, |
| 2636 |
4={label in legend={text=4}}, |
| 2637 |
5={label in legend={text=5}}, |
| 2638 |
6={label in legend={text=6}}, |
| 2639 |
7={label in legend={text=7}}, |
| 2640 |
8={label in legend={text=8}} |
| 2641 |
} |
| 2642 |
}}, |
| 2643 |
] |
| 2644 |
\tikz \datavisualization [ |
| 2645 |
visualize as smooth line/.list={1,2,3,4,5,6,7,8}, |
| 2646 |
legend example, style sheet=vary hue, |
| 2647 |
main legend={max columns=3,ideal number of rows=2}] |
| 2648 |
data group {sin functions}; |
| 2649 |
\end{codeexample} |
| 2650 |
% |
| 2651 |
\end{key} |
| 2652 |
|
| 2653 |
\begin{key}{/tikz/data visualization/legend options/columns=\meta{number}} |
| 2654 |
Shorthand for |ideal number of columns=|\meta{number}. |
| 2655 |
\end{key} |
| 2656 |
|
| 2657 |
|
| 2658 |
\subsubsection{Legend Placement: The General Mechanism} |
| 2659 |
|
| 2660 |
A legend can either be placed next to the data visualization or inside the data |
| 2661 |
visualization at some place where there are no data entries. Both approached |
| 2662 |
have advantages: Placing the legend next to the visualization minimises the |
| 2663 |
``cluttering'' by keeping all the extra information apart from the actual data, |
| 2664 |
while placing the legend inside the visualization minimises the distance |
| 2665 |
between the data sets and their explanations, making it easier for the eye to |
| 2666 |
connect them. |
| 2667 |
|
| 2668 |
For both approaches there are options that make the placement easier, see |
| 2669 |
Sections \ref{section-dv-legend-outside} and~\ref{section-dv-legend-inside}, |
| 2670 |
but these options internally just map to the following two options: |
| 2671 |
|
| 2672 |
\begin{key}{/tikz/data visualization/legend options/anchor=\meta{anchor}} |
| 2673 |
The whole legend is a \tikzname-matrix internally. Thus, in particular, it |
| 2674 |
is stored in a node, which has anchors. Like for any other node, when the |
| 2675 |
node is shown, the node is shifted in such a way that the \meta{anchor} of |
| 2676 |
the node lies at the current |at| position. |
| 2677 |
\end{key} |
| 2678 |
|
| 2679 |
\begin{key}{/tikz/data visualization/legend options/at=\meta{coordinate}} |
| 2680 |
Configures the \meta{coordinate} at which the \meta{anchor} of the legend's |
| 2681 |
node should lie. |
| 2682 |
|
| 2683 |
It may seem hard to predict a good \meta{coordinate} for a legend since, |
| 2684 |
depending of the size of the axis, different positions need to the chosen |
| 2685 |
for the legend. However, it turns out that one can often use the |
| 2686 |
coordinates of the special nodes |data bounding box| and |
| 2687 |
|data visualization bounding box|, documented in |
| 2688 |
Section~\ref{section-dv-bounding-box}. |
| 2689 |
|
| 2690 |
As an example, let us put a legend to the right of the visualization, but |
| 2691 |
so that the first entry starts at the top of the visualization: |
| 2692 |
% |
| 2693 |
\begin{codeexample}[ |
| 2694 |
width=8cm, |
| 2695 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2696 |
pre={\tikz \datavisualization data group {function classes} = { |
| 2697 |
data [set=log, format=function] { |
| 2698 |
var x : interval [0.2:2.5]; |
| 2699 |
func y = ln(\value x); |
| 2700 |
} |
| 2701 |
data [set=lin, format=function] { |
| 2702 |
var x : interval [-2:2.5]; |
| 2703 |
func y = 0.5*\value x; |
| 2704 |
} |
| 2705 |
data [set=squared, format=function] { |
| 2706 |
var x : interval [-1.5:1.5]; |
| 2707 |
func y = \value x*\value x; |
| 2708 |
} |
| 2709 |
data [set=exp, format=function] { |
| 2710 |
var x : interval [-2.5:1]; |
| 2711 |
func y = exp(\value x); |
| 2712 |
} |
| 2713 |
};}, |
| 2714 |
] |
| 2715 |
\tikz \datavisualization [ |
| 2716 |
scientific axes, x axis={label=$x$}, |
| 2717 |
visualize as smooth line/.list= |
| 2718 |
{log, lin, squared, exp}, |
| 2719 |
legend={anchor=north west, at= |
| 2720 |
(data visualization bounding box.north east)}, |
| 2721 |
log= {label in legend={text=$\log x$}}, |
| 2722 |
lin= {label in legend={text=$x/2$}}, |
| 2723 |
squared={label in legend={text=$x^2$}}, |
| 2724 |
exp= {label in legend={text=$e^x$}}, |
| 2725 |
style sheet=vary dashing] |
| 2726 |
data group {function classes}; |
| 2727 |
\end{codeexample} |
| 2728 |
% |
| 2729 |
As can be seen, a bit of an additional shift might have been in order, but |
| 2730 |
the result is otherwise quite satisfactory. |
| 2731 |
\end{key} |
| 2732 |
|
| 2733 |
|
| 2734 |
\subsubsection{Legend Placement: Outside to the Data Visualization} |
| 2735 |
\label{section-dv-legend-outside} |
| 2736 |
|
| 2737 |
The following keys make it easy to place a legend outside the data |
| 2738 |
visualization. |
| 2739 |
|
| 2740 |
\begin{key}{/tikz/data visualization/legend options/east outside} |
| 2741 |
Placing the legend to the right of the data visualization is the default: |
| 2742 |
% |
| 2743 |
\begin{codeexample}[ |
| 2744 |
width=8cm, |
| 2745 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2746 |
pre={\tikz \datavisualization data group {function classes} = { |
| 2747 |
data [set=log, format=function] { |
| 2748 |
var x : interval [0.2:2.5]; |
| 2749 |
func y = ln(\value x); |
| 2750 |
} |
| 2751 |
data [set=lin, format=function] { |
| 2752 |
var x : interval [-2:2.5]; |
| 2753 |
func y = 0.5*\value x; |
| 2754 |
} |
| 2755 |
data [set=squared, format=function] { |
| 2756 |
var x : interval [-1.5:1.5]; |
| 2757 |
func y = \value x*\value x; |
| 2758 |
} |
| 2759 |
data [set=exp, format=function] { |
| 2760 |
var x : interval [-2.5:1]; |
| 2761 |
func y = exp(\value x); |
| 2762 |
} |
| 2763 |
};}, |
| 2764 |
] |
| 2765 |
\tikz \datavisualization [ |
| 2766 |
scientific axes, |
| 2767 |
visualize as smooth line/.list= |
| 2768 |
{log, lin, squared, exp}, |
| 2769 |
legend=east outside, |
| 2770 |
log= {label in legend={text=$\log x$}}, |
| 2771 |
lin= {label in legend={text=$x/2$}}, |
| 2772 |
squared={label in legend={text=$x^2$}}, |
| 2773 |
exp= {label in legend={text=$e^x$}}, |
| 2774 |
style sheet=strong colors] |
| 2775 |
data group {function classes}; |
| 2776 |
\end{codeexample} |
| 2777 |
|
| 2778 |
\begin{key}{/tikz/data visualization/legend options/right} |
| 2779 |
This is an easier-to-remember alias. |
| 2780 |
\end{key} |
| 2781 |
\end{key} |
| 2782 |
|
| 2783 |
\begin{key}{/tikz/data visualization/legend options/north east outside} |
| 2784 |
A variant, where the legend is to the right, but aligned with the northern |
| 2785 |
end of the data visualization: |
| 2786 |
% |
| 2787 |
\begin{codeexample}[ |
| 2788 |
width=8cm, |
| 2789 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2790 |
pre={\tikz \datavisualization data group {function classes} = { |
| 2791 |
data [set=log, format=function] { |
| 2792 |
var x : interval [0.2:2.5]; |
| 2793 |
func y = ln(\value x); |
| 2794 |
} |
| 2795 |
data [set=lin, format=function] { |
| 2796 |
var x : interval [-2:2.5]; |
| 2797 |
func y = 0.5*\value x; |
| 2798 |
} |
| 2799 |
data [set=squared, format=function] { |
| 2800 |
var x : interval [-1.5:1.5]; |
| 2801 |
func y = \value x*\value x; |
| 2802 |
} |
| 2803 |
data [set=exp, format=function] { |
| 2804 |
var x : interval [-2.5:1]; |
| 2805 |
func y = exp(\value x); |
| 2806 |
} |
| 2807 |
};}, |
| 2808 |
] |
| 2809 |
\tikz \datavisualization [ |
| 2810 |
scientific axes, |
| 2811 |
visualize as smooth line/.list= |
| 2812 |
{log, lin, squared, exp}, |
| 2813 |
legend=north east outside, |
| 2814 |
log= {label in legend={text=$\log x$}}, |
| 2815 |
lin= {label in legend={text=$x/2$}}, |
| 2816 |
squared={label in legend={text=$x^2$}}, |
| 2817 |
exp= {label in legend={text=$e^x$}}, |
| 2818 |
style sheet=strong colors] |
| 2819 |
data group {function classes}; |
| 2820 |
\end{codeexample} |
| 2821 |
% |
| 2822 |
\end{key} |
| 2823 |
|
| 2824 |
\begin{key}{/tikz/data visualization/legend options/south east outside} |
| 2825 |
\begin{codeexample}[ |
| 2826 |
width=8cm, |
| 2827 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2828 |
pre={\tikz \datavisualization data group {function classes} = { |
| 2829 |
data [set=log, format=function] { |
| 2830 |
var x : interval [0.2:2.5]; |
| 2831 |
func y = ln(\value x); |
| 2832 |
} |
| 2833 |
data [set=lin, format=function] { |
| 2834 |
var x : interval [-2:2.5]; |
| 2835 |
func y = 0.5*\value x; |
| 2836 |
} |
| 2837 |
data [set=squared, format=function] { |
| 2838 |
var x : interval [-1.5:1.5]; |
| 2839 |
func y = \value x*\value x; |
| 2840 |
} |
| 2841 |
data [set=exp, format=function] { |
| 2842 |
var x : interval [-2.5:1]; |
| 2843 |
func y = exp(\value x); |
| 2844 |
} |
| 2845 |
};}, |
| 2846 |
] |
| 2847 |
\tikz \datavisualization [ |
| 2848 |
scientific axes, |
| 2849 |
visualize as smooth line/.list= |
| 2850 |
{log, lin, squared, exp}, |
| 2851 |
legend=south east outside, |
| 2852 |
log= {label in legend={text=$\log x$}}, |
| 2853 |
lin= {label in legend={text=$x/2$}}, |
| 2854 |
squared={label in legend={text=$x^2$}}, |
| 2855 |
exp= {label in legend={text=$e^x$}}, |
| 2856 |
style sheet=strong colors] |
| 2857 |
data group {function classes}; |
| 2858 |
\end{codeexample} |
| 2859 |
\end{key} |
| 2860 |
|
| 2861 |
\begin{key}{/tikz/data visualization/legend options/west outside} |
| 2862 |
The legend is placed left. Note that the text also swaps its position. |
| 2863 |
% |
| 2864 |
\begin{codeexample}[ |
| 2865 |
width=8cm, |
| 2866 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2867 |
pre={\tikz \datavisualization data group {function classes} = { |
| 2868 |
data [set=log, format=function] { |
| 2869 |
var x : interval [0.2:2.5]; |
| 2870 |
func y = ln(\value x); |
| 2871 |
} |
| 2872 |
data [set=lin, format=function] { |
| 2873 |
var x : interval [-2:2.5]; |
| 2874 |
func y = 0.5*\value x; |
| 2875 |
} |
| 2876 |
data [set=squared, format=function] { |
| 2877 |
var x : interval [-1.5:1.5]; |
| 2878 |
func y = \value x*\value x; |
| 2879 |
} |
| 2880 |
data [set=exp, format=function] { |
| 2881 |
var x : interval [-2.5:1]; |
| 2882 |
func y = exp(\value x); |
| 2883 |
} |
| 2884 |
};}, |
| 2885 |
] |
| 2886 |
\tikz \datavisualization [ |
| 2887 |
scientific axes, |
| 2888 |
visualize as smooth line/.list= |
| 2889 |
{log, lin, squared, exp}, |
| 2890 |
legend=west outside, |
| 2891 |
log= {label in legend={text=$\log x$}}, |
| 2892 |
lin= {label in legend={text=$x/2$}}, |
| 2893 |
squared={label in legend={text=$x^2$}}, |
| 2894 |
exp= {label in legend={text=$e^x$}}, |
| 2895 |
style sheet=strong colors] |
| 2896 |
data group {function classes}; |
| 2897 |
\end{codeexample} |
| 2898 |
% |
| 2899 |
\begin{key}{/tikz/data visualization/legend options/left} |
| 2900 |
This is an easier-to-remember alias. |
| 2901 |
\end{key} |
| 2902 |
\end{key} |
| 2903 |
|
| 2904 |
\begin{key}{/tikz/data visualization/legend options/north west outside} |
| 2905 |
\begin{codeexample}[ |
| 2906 |
width=8cm, |
| 2907 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2908 |
pre={\tikz \datavisualization data group {function classes} = { |
| 2909 |
data [set=log, format=function] { |
| 2910 |
var x : interval [0.2:2.5]; |
| 2911 |
func y = ln(\value x); |
| 2912 |
} |
| 2913 |
data [set=lin, format=function] { |
| 2914 |
var x : interval [-2:2.5]; |
| 2915 |
func y = 0.5*\value x; |
| 2916 |
} |
| 2917 |
data [set=squared, format=function] { |
| 2918 |
var x : interval [-1.5:1.5]; |
| 2919 |
func y = \value x*\value x; |
| 2920 |
} |
| 2921 |
data [set=exp, format=function] { |
| 2922 |
var x : interval [-2.5:1]; |
| 2923 |
func y = exp(\value x); |
| 2924 |
} |
| 2925 |
};}, |
| 2926 |
] |
| 2927 |
\tikz \datavisualization [ |
| 2928 |
scientific axes, |
| 2929 |
visualize as smooth line/.list= |
| 2930 |
{log, lin, squared, exp}, |
| 2931 |
legend=north west outside, |
| 2932 |
log= {label in legend={text=$\log x$}}, |
| 2933 |
lin= {label in legend={text=$x/2$}}, |
| 2934 |
squared={label in legend={text=$x^2$}}, |
| 2935 |
exp= {label in legend={text=$e^x$}}, |
| 2936 |
style sheet=strong colors] |
| 2937 |
data group {function classes}; |
| 2938 |
\end{codeexample} |
| 2939 |
\end{key} |
| 2940 |
|
| 2941 |
\begin{key}{/tikz/data visualization/legend options/south west outside} |
| 2942 |
\begin{codeexample}[ |
| 2943 |
width=8cm, |
| 2944 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2945 |
pre={\tikz \datavisualization data group {function classes} = { |
| 2946 |
data [set=log, format=function] { |
| 2947 |
var x : interval [0.2:2.5]; |
| 2948 |
func y = ln(\value x); |
| 2949 |
} |
| 2950 |
data [set=lin, format=function] { |
| 2951 |
var x : interval [-2:2.5]; |
| 2952 |
func y = 0.5*\value x; |
| 2953 |
} |
| 2954 |
data [set=squared, format=function] { |
| 2955 |
var x : interval [-1.5:1.5]; |
| 2956 |
func y = \value x*\value x; |
| 2957 |
} |
| 2958 |
data [set=exp, format=function] { |
| 2959 |
var x : interval [-2.5:1]; |
| 2960 |
func y = exp(\value x); |
| 2961 |
} |
| 2962 |
};}, |
| 2963 |
] |
| 2964 |
\tikz \datavisualization [ |
| 2965 |
scientific axes, |
| 2966 |
visualize as smooth line/.list= |
| 2967 |
{log, lin, squared, exp}, |
| 2968 |
legend=south west outside, |
| 2969 |
log= {label in legend={text=$\log x$}}, |
| 2970 |
lin= {label in legend={text=$x/2$}}, |
| 2971 |
squared={label in legend={text=$x^2$}}, |
| 2972 |
exp= {label in legend={text=$e^x$}}, |
| 2973 |
style sheet=strong colors] |
| 2974 |
data group {function classes}; |
| 2975 |
\end{codeexample} |
| 2976 |
\end{key} |
| 2977 |
|
| 2978 |
|
| 2979 |
\begin{key}{/tikz/data visualization/legend options/north outside} |
| 2980 |
The legend is placed above the data. Note that the legend entries now for a |
| 2981 |
row rather than a column. |
| 2982 |
% |
| 2983 |
\begin{codeexample}[ |
| 2984 |
width=8cm, |
| 2985 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 2986 |
pre={\tikz \datavisualization data group {function classes} = { |
| 2987 |
data [set=log, format=function] { |
| 2988 |
var x : interval [0.2:2.5]; |
| 2989 |
func y = ln(\value x); |
| 2990 |
} |
| 2991 |
data [set=lin, format=function] { |
| 2992 |
var x : interval [-2:2.5]; |
| 2993 |
func y = 0.5*\value x; |
| 2994 |
} |
| 2995 |
data [set=squared, format=function] { |
| 2996 |
var x : interval [-1.5:1.5]; |
| 2997 |
func y = \value x*\value x; |
| 2998 |
} |
| 2999 |
data [set=exp, format=function] { |
| 3000 |
var x : interval [-2.5:1]; |
| 3001 |
func y = exp(\value x); |
| 3002 |
} |
| 3003 |
};}, |
| 3004 |
] |
| 3005 |
\tikz \datavisualization [ |
| 3006 |
scientific axes, |
| 3007 |
visualize as smooth line/.list= |
| 3008 |
{log, lin, squared, exp}, |
| 3009 |
legend=north outside, |
| 3010 |
log= {label in legend={text=$\log x$}}, |
| 3011 |
lin= {label in legend={text=$x/2$}}, |
| 3012 |
squared={label in legend={text=$x^2$}}, |
| 3013 |
exp= {label in legend={text=$e^x$}}, |
| 3014 |
style sheet=strong colors] |
| 3015 |
data group {function classes}; |
| 3016 |
\end{codeexample} |
| 3017 |
% |
| 3018 |
\begin{key}{/tikz/data visualization/legend options/above} |
| 3019 |
This is an easier-to-remember alias. |
| 3020 |
\end{key} |
| 3021 |
\end{key} |
| 3022 |
|
| 3023 |
\begin{key}{/tikz/data visualization/legend options/south outside} |
| 3024 |
\begin{codeexample}[ |
| 3025 |
width=8cm, |
| 3026 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 3027 |
pre={\tikz \datavisualization data group {function classes} = { |
| 3028 |
data [set=log, format=function] { |
| 3029 |
var x : interval [0.2:2.5]; |
| 3030 |
func y = ln(\value x); |
| 3031 |
} |
| 3032 |
data [set=lin, format=function] { |
| 3033 |
var x : interval [-2:2.5]; |
| 3034 |
func y = 0.5*\value x; |
| 3035 |
} |
| 3036 |
data [set=squared, format=function] { |
| 3037 |
var x : interval [-1.5:1.5]; |
| 3038 |
func y = \value x*\value x; |
| 3039 |
} |
| 3040 |
data [set=exp, format=function] { |
| 3041 |
var x : interval [-2.5:1]; |
| 3042 |
func y = exp(\value x); |
| 3043 |
} |
| 3044 |
};}, |
| 3045 |
] |
| 3046 |
\tikz \datavisualization [ |
| 3047 |
scientific axes, |
| 3048 |
visualize as smooth line/.list= |
| 3049 |
{log, lin, squared, exp}, |
| 3050 |
legend=south outside, |
| 3051 |
log= {label in legend={text=$\log x$}}, |
| 3052 |
lin= {label in legend={text=$x/2$}}, |
| 3053 |
squared={label in legend={text=$x^2$}}, |
| 3054 |
exp= {label in legend={text=$e^x$}}, |
| 3055 |
style sheet=strong colors] |
| 3056 |
data group {function classes}; |
| 3057 |
\end{codeexample} |
| 3058 |
% |
| 3059 |
\begin{key}{/tikz/data visualization/legend options/below} |
| 3060 |
This is an easier-to-remember alias. |
| 3061 |
\end{key} |
| 3062 |
\end{key} |
| 3063 |
|
| 3064 |
|
| 3065 |
\subsubsection{Legend Placement: Inside to the Data Visualization} |
| 3066 |
\label{section-dv-legend-inside} |
| 3067 |
|
| 3068 |
There are two sets of options for placing a legend directly inside a data |
| 3069 |
visualization: First, there are options for placing it inside, but next to some |
| 3070 |
part of the border. Second, there are options for positioning it relative to a |
| 3071 |
coordinate given by a certain data point. |
| 3072 |
|
| 3073 |
\begin{key}{/tikz/data visualization/legend options/south east inside} |
| 3074 |
Puts the legend in the upper right corner of the data. |
| 3075 |
% |
| 3076 |
\begin{codeexample}[ |
| 3077 |
width=8cm, |
| 3078 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 3079 |
pre={\tikz \datavisualization data group {function classes} = { |
| 3080 |
data [set=log, format=function] { |
| 3081 |
var x : interval [0.2:2.5]; |
| 3082 |
func y = ln(\value x); |
| 3083 |
} |
| 3084 |
data [set=lin, format=function] { |
| 3085 |
var x : interval [-2:2.5]; |
| 3086 |
func y = 0.5*\value x; |
| 3087 |
} |
| 3088 |
data [set=squared, format=function] { |
| 3089 |
var x : interval [-1.5:1.5]; |
| 3090 |
func y = \value x*\value x; |
| 3091 |
} |
| 3092 |
data [set=exp, format=function] { |
| 3093 |
var x : interval [-2.5:1]; |
| 3094 |
func y = exp(\value x); |
| 3095 |
} |
| 3096 |
};}, |
| 3097 |
] |
| 3098 |
\tikz \datavisualization [ |
| 3099 |
scientific axes, |
| 3100 |
visualize as smooth line/.list= |
| 3101 |
{log, lin}, |
| 3102 |
legend=south east inside, |
| 3103 |
log= {label in legend={text=$\log x$}}, |
| 3104 |
lin= {label in legend={text=$x/2$}}, |
| 3105 |
style sheet=strong colors] |
| 3106 |
data group {function classes}; |
| 3107 |
\end{codeexample} |
| 3108 |
|
| 3109 |
Note that the text is now a little smaller since there tends to be much |
| 3110 |
less space inside the data visualization than next to it. Also, the |
| 3111 |
legend's node is filled in white by default to ensures that the legend is |
| 3112 |
clearly legible even in the presence of, say, a grid or data points behind |
| 3113 |
it. This behavior is triggered by the following style key: |
| 3114 |
|
| 3115 |
\begin{stylekey}{/tikz/data visualization/legend options/every legend inside} |
| 3116 |
Executed the keys |opaque| by default and sets the text size to the |
| 3117 |
size of footnotes. |
| 3118 |
\end{stylekey} |
| 3119 |
\end{key} |
| 3120 |
|
| 3121 |
In order to further configure the default appearance of an inner legend, the |
| 3122 |
following keys might be useful: |
| 3123 |
|
| 3124 |
\begin{key}{/tikz/data visualization/legend options/opaque=\meta{color} (default white)} |
| 3125 |
When this key is used, the legend's node will be filled with the |
| 3126 |
\meta{color} and its corners will be rounded. Additionally, the inner and |
| 3127 |
outer separations will be set to sensible values. |
| 3128 |
\end{key} |
| 3129 |
% |
| 3130 |
\begin{key}{/tikz/data visualization/legend options/transparent} |
| 3131 |
Sets the filling of the legend node to |none|. |
| 3132 |
\end{key} |
| 3133 |
|
| 3134 |
The following keys work much the same way as |south east inside|: |
| 3135 |
|
| 3136 |
\begin{key}{/tikz/data visualization/legend options/east inside} |
| 3137 |
\end{key} |
| 3138 |
% |
| 3139 |
\begin{key}{/tikz/data visualization/legend options/north east inside} |
| 3140 |
\end{key} |
| 3141 |
% |
| 3142 |
\begin{key}{/tikz/data visualization/legend options/south west inside} |
| 3143 |
\end{key} |
| 3144 |
% |
| 3145 |
\begin{key}{/tikz/data visualization/legend options/west inside} |
| 3146 |
\end{key} |
| 3147 |
% |
| 3148 |
\begin{key}{/tikz/data visualization/legend options/north west inside} |
| 3149 |
\end{key} |
| 3150 |
|
| 3151 |
The keys |south inside| and |north inside| are a bit different: They use a row |
| 3152 |
rather than a column for the legend entries: |
| 3153 |
|
| 3154 |
\begin{key}{/tikz/data visualization/legend options/south inside} |
| 3155 |
Puts the legend in the upper right corner of the data. Note that the text |
| 3156 |
is now a little smaller since there tends to be much less space inside the |
| 3157 |
data visualization than next to it. |
| 3158 |
% |
| 3159 |
\begin{codeexample}[ |
| 3160 |
width=8cm, |
| 3161 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 3162 |
pre={\tikz \datavisualization data group {function classes} = { |
| 3163 |
data [set=log, format=function] { |
| 3164 |
var x : interval [0.2:2.5]; |
| 3165 |
func y = ln(\value x); |
| 3166 |
} |
| 3167 |
data [set=lin, format=function] { |
| 3168 |
var x : interval [-2:2.5]; |
| 3169 |
func y = 0.5*\value x; |
| 3170 |
} |
| 3171 |
data [set=squared, format=function] { |
| 3172 |
var x : interval [-1.5:1.5]; |
| 3173 |
func y = \value x*\value x; |
| 3174 |
} |
| 3175 |
data [set=exp, format=function] { |
| 3176 |
var x : interval [-2.5:1]; |
| 3177 |
func y = exp(\value x); |
| 3178 |
} |
| 3179 |
};}, |
| 3180 |
] |
| 3181 |
\tikz \datavisualization [ |
| 3182 |
scientific axes, |
| 3183 |
visualize as smooth line/.list={log, lin}, |
| 3184 |
legend=south inside, |
| 3185 |
log= {label in legend={text=$\log x$}}, |
| 3186 |
lin= {label in legend={text=$x/2$}}, |
| 3187 |
style sheet=strong colors] |
| 3188 |
data group {function classes}; |
| 3189 |
\end{codeexample} |
| 3190 |
% |
| 3191 |
\end{key} |
| 3192 |
|
| 3193 |
\begin{key}{/tikz/data visualization/legend options/north inside} |
| 3194 |
As above. |
| 3195 |
\end{key} |
| 3196 |
|
| 3197 |
The above keys do not always give you as fine a control as you may need over |
| 3198 |
the placement of the legend. In such cases, the following keys may help (or you |
| 3199 |
can revert to directly setting the |at| and the |anchor| keys): |
| 3200 |
|
| 3201 |
\begin{key}{/tikz/data visualization/legend options/at values=\meta{data point}} |
| 3202 |
This key allows you to specify the desired center of the legend in terms of |
| 3203 |
a data point. The \meta{data point} should be a list of comma-separated |
| 3204 |
key--value pairs that specify a data point. The legend will then be |
| 3205 |
centered at this data point. |
| 3206 |
% |
| 3207 |
\begin{codeexample}[ |
| 3208 |
width=6cm, |
| 3209 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 3210 |
pre={\tikz \datavisualization data group {function classes} = { |
| 3211 |
data [set=log, format=function] { |
| 3212 |
var x : interval [0.2:2.5]; |
| 3213 |
func y = ln(\value x); |
| 3214 |
} |
| 3215 |
data [set=lin, format=function] { |
| 3216 |
var x : interval [-2:2.5]; |
| 3217 |
func y = 0.5*\value x; |
| 3218 |
} |
| 3219 |
data [set=squared, format=function] { |
| 3220 |
var x : interval [-1.5:1.5]; |
| 3221 |
func y = \value x*\value x; |
| 3222 |
} |
| 3223 |
data [set=exp, format=function] { |
| 3224 |
var x : interval [-2.5:1]; |
| 3225 |
func y = exp(\value x); |
| 3226 |
} |
| 3227 |
};}, |
| 3228 |
] |
| 3229 |
\tikz \datavisualization [ |
| 3230 |
scientific axes, |
| 3231 |
visualize as smooth line/.list={log, lin}, |
| 3232 |
legend={at values={x=-1, y=2}}, |
| 3233 |
log= {label in legend={text=$\log x$}}, |
| 3234 |
lin= {label in legend={text=$x/2$}}, |
| 3235 |
style sheet=strong colors] |
| 3236 |
data group {function classes}; |
| 3237 |
\end{codeexample} |
| 3238 |
% |
| 3239 |
\end{key} |
| 3240 |
|
| 3241 |
\begin{key}{/tikz/data visualization/legend options/right of=\meta{data point}} |
| 3242 |
Works like |at values|, but the anchor is set to |west|: |
| 3243 |
% |
| 3244 |
\begin{codeexample}[ |
| 3245 |
width=6cm, |
| 3246 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 3247 |
pre={\tikz \datavisualization data group {function classes} = { |
| 3248 |
data [set=log, format=function] { |
| 3249 |
var x : interval [0.2:2.5]; |
| 3250 |
func y = ln(\value x); |
| 3251 |
} |
| 3252 |
data [set=lin, format=function] { |
| 3253 |
var x : interval [-2:2.5]; |
| 3254 |
func y = 0.5*\value x; |
| 3255 |
} |
| 3256 |
data [set=squared, format=function] { |
| 3257 |
var x : interval [-1.5:1.5]; |
| 3258 |
func y = \value x*\value x; |
| 3259 |
} |
| 3260 |
data [set=exp, format=function] { |
| 3261 |
var x : interval [-2.5:1]; |
| 3262 |
func y = exp(\value x); |
| 3263 |
} |
| 3264 |
};}, |
| 3265 |
] |
| 3266 |
\tikz \datavisualization [ |
| 3267 |
scientific axes, |
| 3268 |
visualize as smooth line/.list={log, lin}, |
| 3269 |
legend={right of={x=-1, y=2}}, |
| 3270 |
log= {label in legend={text=$\log x$}}, |
| 3271 |
lin= {label in legend={text=$x/2$}}, |
| 3272 |
style sheet=strong colors] |
| 3273 |
data group {function classes}; |
| 3274 |
\end{codeexample} |
| 3275 |
% |
| 3276 |
\end{key} |
| 3277 |
|
| 3278 |
The following keys work similarly: |
| 3279 |
% |
| 3280 |
\begin{key}{/tikz/data visualization/legend options/above right of=\meta{data point}} |
| 3281 |
\end{key} |
| 3282 |
% |
| 3283 |
\begin{key}{/tikz/data visualization/legend options/above of=\meta{data point}} |
| 3284 |
\end{key} |
| 3285 |
% |
| 3286 |
\begin{key}{/tikz/data visualization/legend options/above left of=\meta{data point}} |
| 3287 |
\end{key} |
| 3288 |
% |
| 3289 |
\begin{key}{/tikz/data visualization/legend options/left of=\meta{data point}} |
| 3290 |
\end{key} |
| 3291 |
% |
| 3292 |
\begin{key}{/tikz/data visualization/legend options/below left of=\meta{data point}} |
| 3293 |
\end{key} |
| 3294 |
% |
| 3295 |
\begin{key}{/tikz/data visualization/legend options/below of=\meta{data point}} |
| 3296 |
\end{key} |
| 3297 |
% |
| 3298 |
\begin{key}{/tikz/data visualization/legend options/below right of=\meta{data point}} |
| 3299 |
\end{key} |
| 3300 |
|
| 3301 |
|
| 3302 |
\subsubsection{Legend Entries: General Styling} |
| 3303 |
\label{section-dv-label-legend-entry-options} |
| 3304 |
|
| 3305 |
The entries in a legend can be styled in several ways: |
| 3306 |
% |
| 3307 |
\begin{itemize} |
| 3308 |
\item You can configure the styling of the text node. |
| 3309 |
\item You can configure the relative placement of the text node and the |
| 3310 |
little picture depicting the data set's styling. |
| 3311 |
\item You can configure how the data set's styling is depicted. |
| 3312 |
\end{itemize} |
| 3313 |
|
| 3314 |
Before we have look at how each of these are configured, in detail, let us |
| 3315 |
first have a look at the keys that allow us to save a set of such styles: |
| 3316 |
|
| 3317 |
\begin{stylekey}{/tikz/data visualization/every label in legend} |
| 3318 |
This key is executed with every label in a legend. However, the options |
| 3319 |
stored in this style are executed with the path prefix |
| 3320 |
|/tikz/data visualization/legend entry options|. Thus, this key can use |
| 3321 |
keys like |node style| to configure the styling of all text nodes: |
| 3322 |
% |
| 3323 |
\begin{codeexample}[ |
| 3324 |
width=8cm, |
| 3325 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 3326 |
pre={\tikz \datavisualization data group {function classes} = { |
| 3327 |
data [set=log, format=function] { |
| 3328 |
var x : interval [0.2:2.5]; |
| 3329 |
func y = ln(\value x); |
| 3330 |
} |
| 3331 |
data [set=lin, format=function] { |
| 3332 |
var x : interval [-2:2.5]; |
| 3333 |
func y = 0.5*\value x; |
| 3334 |
} |
| 3335 |
data [set=squared, format=function] { |
| 3336 |
var x : interval [-1.5:1.5]; |
| 3337 |
func y = \value x*\value x; |
| 3338 |
} |
| 3339 |
data [set=exp, format=function] { |
| 3340 |
var x : interval [-2.5:1]; |
| 3341 |
func y = exp(\value x); |
| 3342 |
} |
| 3343 |
};}, |
| 3344 |
] |
| 3345 |
\tikz \datavisualization [ |
| 3346 |
scientific axes, |
| 3347 |
every label in legend/.style={node style= |
| 3348 |
{fill=red!30}}, |
| 3349 |
visualize as smooth line/.list= |
| 3350 |
{log, lin, squared, exp}, |
| 3351 |
legend=north east outside, |
| 3352 |
log= {label in legend={text=$\log x$}}, |
| 3353 |
lin= {label in legend={text=$x/2$, |
| 3354 |
node style={circle, draw=red}}}, |
| 3355 |
squared={label in legend={text=$x^2$}}, |
| 3356 |
exp= {label in legend={text=$e^x$}}, |
| 3357 |
style sheet=strong colors] |
| 3358 |
data group {function classes}; |
| 3359 |
\end{codeexample} |
| 3360 |
% |
| 3361 |
\end{stylekey} |
| 3362 |
|
| 3363 |
\begin{key}{/tikz/data visualization/legend options/label style=\meta{options}} |
| 3364 |
This key can be used with a legend. It will simply add the \meta{options} |
| 3365 |
to the |every label in legend| style for the given legend. |
| 3366 |
% |
| 3367 |
\begin{codeexample}[ |
| 3368 |
width=8cm, |
| 3369 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 3370 |
pre={\tikz \datavisualization data group {function classes} = { |
| 3371 |
data [set=log, format=function] { |
| 3372 |
var x : interval [0.2:2.5]; |
| 3373 |
func y = ln(\value x); |
| 3374 |
} |
| 3375 |
data [set=lin, format=function] { |
| 3376 |
var x : interval [-2:2.5]; |
| 3377 |
func y = 0.5*\value x; |
| 3378 |
} |
| 3379 |
data [set=squared, format=function] { |
| 3380 |
var x : interval [-1.5:1.5]; |
| 3381 |
func y = \value x*\value x; |
| 3382 |
} |
| 3383 |
data [set=exp, format=function] { |
| 3384 |
var x : interval [-2.5:1]; |
| 3385 |
func y = exp(\value x); |
| 3386 |
} |
| 3387 |
};}, |
| 3388 |
] |
| 3389 |
\tikz \datavisualization [ |
| 3390 |
scientific axes, |
| 3391 |
visualize as smooth line/.list= |
| 3392 |
{log, lin, squared, exp}, |
| 3393 |
legend={label style={node style=draw}}, |
| 3394 |
log= {label in legend={text=$\log x$}}, |
| 3395 |
lin= {label in legend={text=$x/2$, |
| 3396 |
node style={circle, draw=red}}}, |
| 3397 |
squared={label in legend={text=$x^2$}}, |
| 3398 |
exp= {label in legend={text=$e^x$}}, |
| 3399 |
style sheet=strong colors] |
| 3400 |
data group {function classes}; |
| 3401 |
\end{codeexample} |
| 3402 |
% |
| 3403 |
\end{key} |
| 3404 |
|
| 3405 |
|
| 3406 |
\subsubsection{Legend Entries: Styling the Text Node} |
| 3407 |
|
| 3408 |
The appearance of the text nodes is easy to configure. |
| 3409 |
|
| 3410 |
\begin{key}{/tikz/data visualization/legend entry options/node style=\meta{options}} |
| 3411 |
This key adds \meta{options} to the styling of the text nodes of the label. |
| 3412 |
% |
| 3413 |
\begin{codeexample}[ |
| 3414 |
width=8cm, |
| 3415 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 3416 |
pre={\tikz \datavisualization data group {function classes} = { |
| 3417 |
data [set=log, format=function] { |
| 3418 |
var x : interval [0.2:2.5]; |
| 3419 |
func y = ln(\value x); |
| 3420 |
} |
| 3421 |
data [set=lin, format=function] { |
| 3422 |
var x : interval [-2:2.5]; |
| 3423 |
func y = 0.5*\value x; |
| 3424 |
} |
| 3425 |
data [set=squared, format=function] { |
| 3426 |
var x : interval [-1.5:1.5]; |
| 3427 |
func y = \value x*\value x; |
| 3428 |
} |
| 3429 |
data [set=exp, format=function] { |
| 3430 |
var x : interval [-2.5:1]; |
| 3431 |
func y = exp(\value x); |
| 3432 |
} |
| 3433 |
};}, |
| 3434 |
] |
| 3435 |
\tikz \datavisualization [ |
| 3436 |
scientific axes, |
| 3437 |
visualize as smooth line/.list= |
| 3438 |
{log, lin, squared, exp}, |
| 3439 |
legend=north east outside, |
| 3440 |
log= {label in legend={text=$\log x$}}, |
| 3441 |
lin= {label in legend={text=$x/2$, |
| 3442 |
node style={circle, draw=red}}}, |
| 3443 |
squared={label in legend={text=$x^2$}}, |
| 3444 |
exp= {label in legend={text=$e^x$}}, |
| 3445 |
style sheet=strong colors] |
| 3446 |
data group {function classes}; |
| 3447 |
\end{codeexample} |
| 3448 |
% |
| 3449 |
\end{key} |
| 3450 |
|
| 3451 |
\begin{key}{/tikz/data visualization/legend entry options/text colored} |
| 3452 |
Causes the |node style| to set the text color to |visualizer color|. The |
| 3453 |
effect of this is that the label's text will have the same color as the |
| 3454 |
data set to which it is attached. |
| 3455 |
% |
| 3456 |
\begin{codeexample}[ |
| 3457 |
width=8cm, |
| 3458 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 3459 |
pre={\tikz \datavisualization data group {function classes} = { |
| 3460 |
data [set=log, format=function] { |
| 3461 |
var x : interval [0.2:2.5]; |
| 3462 |
func y = ln(\value x); |
| 3463 |
} |
| 3464 |
data [set=lin, format=function] { |
| 3465 |
var x : interval [-2:2.5]; |
| 3466 |
func y = 0.5*\value x; |
| 3467 |
} |
| 3468 |
data [set=squared, format=function] { |
| 3469 |
var x : interval [-1.5:1.5]; |
| 3470 |
func y = \value x*\value x; |
| 3471 |
} |
| 3472 |
data [set=exp, format=function] { |
| 3473 |
var x : interval [-2.5:1]; |
| 3474 |
func y = exp(\value x); |
| 3475 |
} |
| 3476 |
};}, |
| 3477 |
] |
| 3478 |
\tikz \datavisualization [ |
| 3479 |
scientific axes, |
| 3480 |
visualize as smooth line/.list= |
| 3481 |
{log, lin, squared, exp}, |
| 3482 |
legend={label style=text colored}, |
| 3483 |
log= {label in legend={text=$\log x$}}, |
| 3484 |
lin= {label in legend={text=$x/2$}}, |
| 3485 |
squared={label in legend={text=$x^2$}}, |
| 3486 |
exp= {label in legend={text=$e^x$}}, |
| 3487 |
style sheet=strong colors] |
| 3488 |
data group {function classes}; |
| 3489 |
\end{codeexample} |
| 3490 |
% |
| 3491 |
\end{key} |
| 3492 |
|
| 3493 |
|
| 3494 |
\subsubsection{Legend Entries: Text Placement} |
| 3495 |
|
| 3496 |
Three keys govern where the text will be placed relative to the data set style |
| 3497 |
visualization. |
| 3498 |
|
| 3499 |
\begin{key}{/tikz/data visualization/legend entry options/text right} |
| 3500 |
Placed the text node to the right of the data set style visualization. This |
| 3501 |
is the default for most, but not all, legends. |
| 3502 |
\end{key} |
| 3503 |
% |
| 3504 |
\begin{key}{/tikz/data visualization/legend entry options/text left} |
| 3505 |
Placed the text node to the left of the data set style visualization. |
| 3506 |
% |
| 3507 |
\begin{codeexample}[ |
| 3508 |
width=8cm, |
| 3509 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 3510 |
pre={\tikz \datavisualization data group {function classes} = { |
| 3511 |
data [set=log, format=function] { |
| 3512 |
var x : interval [0.2:2.5]; |
| 3513 |
func y = ln(\value x); |
| 3514 |
} |
| 3515 |
data [set=lin, format=function] { |
| 3516 |
var x : interval [-2:2.5]; |
| 3517 |
func y = 0.5*\value x; |
| 3518 |
} |
| 3519 |
data [set=squared, format=function] { |
| 3520 |
var x : interval [-1.5:1.5]; |
| 3521 |
func y = \value x*\value x; |
| 3522 |
} |
| 3523 |
data [set=exp, format=function] { |
| 3524 |
var x : interval [-2.5:1]; |
| 3525 |
func y = exp(\value x); |
| 3526 |
} |
| 3527 |
};}, |
| 3528 |
] |
| 3529 |
\tikz \datavisualization [ |
| 3530 |
scientific axes, |
| 3531 |
visualize as smooth line/.list= |
| 3532 |
{log, lin, squared, exp}, |
| 3533 |
legend={label style=text left}, |
| 3534 |
log= {label in legend={text=$\log x$}}, |
| 3535 |
lin= {label in legend={text=$x/2$}}, |
| 3536 |
squared={label in legend={text=$x^2$}}, |
| 3537 |
exp= {label in legend={text=$e^x$}}, |
| 3538 |
style sheet=strong colors] |
| 3539 |
data group {function classes}; |
| 3540 |
\end{codeexample} |
| 3541 |
% |
| 3542 |
\end{key} |
| 3543 |
% |
| 3544 |
\begin{key}{/tikz/data visualization/legend entry options/text only} |
| 3545 |
Shows only the text nodes and no data set style visualization at all. This |
| 3546 |
options only makes sense in conjunction with the |text colored| options, |
| 3547 |
which is why this options is also selected implicitly. |
| 3548 |
% |
| 3549 |
\begin{codeexample}[ |
| 3550 |
width=8cm, |
| 3551 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 3552 |
pre={\tikz \datavisualization data group {function classes} = { |
| 3553 |
data [set=log, format=function] { |
| 3554 |
var x : interval [0.2:2.5]; |
| 3555 |
func y = ln(\value x); |
| 3556 |
} |
| 3557 |
data [set=lin, format=function] { |
| 3558 |
var x : interval [-2:2.5]; |
| 3559 |
func y = 0.5*\value x; |
| 3560 |
} |
| 3561 |
data [set=squared, format=function] { |
| 3562 |
var x : interval [-1.5:1.5]; |
| 3563 |
func y = \value x*\value x; |
| 3564 |
} |
| 3565 |
data [set=exp, format=function] { |
| 3566 |
var x : interval [-2.5:1]; |
| 3567 |
func y = exp(\value x); |
| 3568 |
} |
| 3569 |
};}, |
| 3570 |
] |
| 3571 |
\tikz \datavisualization [ |
| 3572 |
scientific axes, |
| 3573 |
visualize as smooth line/.list= |
| 3574 |
{log, lin, squared, exp}, |
| 3575 |
legend={south east inside, rows=2, |
| 3576 |
label style=text only}, |
| 3577 |
log= {label in legend={text=$\log x$}}, |
| 3578 |
lin= {label in legend={text=$x/2$}}, |
| 3579 |
squared={label in legend={text=$x^2$}}, |
| 3580 |
exp= {label in legend={text=$e^x$}}, |
| 3581 |
style sheet=strong colors] |
| 3582 |
data group {function classes}; |
| 3583 |
\end{codeexample} |
| 3584 |
% |
| 3585 |
\end{key} |
| 3586 |
|
| 3587 |
|
| 3588 |
\subsubsection{Advanced: Labels in Legends and Their Visualizers} |
| 3589 |
\label{section-dv-legend-entries} |
| 3590 |
|
| 3591 |
The following explanations are important only for you if you intend to create a |
| 3592 |
new visualizer and an accompanying label in legend visualizer; otherwise you |
| 3593 |
can safely proceed with the next section. |
| 3594 |
|
| 3595 |
A legend entry consists not only of some explaining text, but, even more |
| 3596 |
importantly, of a visual representation of the style used for the data points, |
| 3597 |
created by a \emph{label in legend visualizer}. For instance, when data points |
| 3598 |
are visualized as lines in different colors, the legend entry for the first |
| 3599 |
line might consist of the text ``first experiment'' and a short line in black |
| 3600 |
and the second entry might consist of ``failed experiment'' and a short line in |
| 3601 |
red -- assuming, of course, that the style sheet makes the first line black and |
| 3602 |
the second line blue. As another example, when data sets are visualized as |
| 3603 |
clouds of plot marks, the texts in the legend would be accompanied by the plot |
| 3604 |
marks used to visualize the data sets. |
| 3605 |
|
| 3606 |
For every visualizer, the \emph{label in legend visualizer} creates an |
| 3607 |
appropriate visualization of the data set's styling. There may be more than one |
| 3608 |
possible such label in legend visualizer that is appropriate, in which case |
| 3609 |
options are used to choose between them. |
| 3610 |
|
| 3611 |
Let us start with the key for creating a new legend entry. This key gets called |
| 3612 |
for instance by |label in legend|: |
| 3613 |
|
| 3614 |
\begin{key}{/tikz/data visualization/new legend entry=\meta{options}} |
| 3615 |
This key will add a new entry to the legend that is identified by the |
| 3616 |
\meta{options}. For this, the \meta{options} are executed once with the |
| 3617 |
path prefix |/tikz/data visualization/legend entry options| and the |
| 3618 |
resulting setting of the |legend| key is used to pick which legend the new |
| 3619 |
entry should belong to. Then, the \meta{options} are stored away for the |
| 3620 |
time being. |
| 3621 |
|
| 3622 |
Later, when the legend is created, the \meta{options} get executed once |
| 3623 |
more. This time, however, the |legend| key is no longer important. Instead, |
| 3624 |
the \meta{options} that setup keys like |text| or |visualizer in legend| |
| 3625 |
now play a role. |
| 3626 |
|
| 3627 |
In detail, the following happens: |
| 3628 |
% |
| 3629 |
\begin{itemize} |
| 3630 |
\item For the legend entry, a little cell picture is created in the |
| 3631 |
matrix of the legend (see Section~\ref{section-tikz-cell-pictures} |
| 3632 |
for details on cell pictures). |
| 3633 |
\item Inside this picture, a node is created whose text is taken from |
| 3634 |
the key |
| 3635 |
% |
| 3636 |
\begin{codeexample}[code only] |
| 3637 |
/tikz/data visualization/legend entry options/text |
| 3638 |
\end{codeexample} |
| 3639 |
% |
| 3640 |
\item Also inside the picture, the code stored in the following key |
| 3641 |
gets executed: |
| 3642 |
% |
| 3643 |
\begin{key}{/tikz/data visualization/legend entry options/visualizer in legend} |
| 3644 |
Set this key to some code that paints something in the cell |
| 3645 |
picture. Typically, this will be a visual representation of the |
| 3646 |
data set styling, but it could also be something different. |
| 3647 |
% |
| 3648 |
\begin{codeexample}[width=6cm,preamble={\usetikzlibrary{datavisualization}}] |
| 3649 |
\tikz \datavisualization [ |
| 3650 |
school book axes, visualize as line/.list={a,b}, |
| 3651 |
style sheet=vary dashing, |
| 3652 |
a={label in legend={text=a}}, |
| 3653 |
new legend entry={ |
| 3654 |
text=spacer, |
| 3655 |
visualizer in legend={\draw[solid] (0,0) circle[radius=2pt];} |
| 3656 |
}, |
| 3657 |
b={label in legend={text=b}}] |
| 3658 |
data point [x=-1, y=-1, set=a] data point [x=1, y=0, set=a] |
| 3659 |
data point [x=-1, y=1, set=b] data point [x=1, y=0.5, set=b]; |
| 3660 |
\end{codeexample} |
| 3661 |
\end{key} |
| 3662 |
\end{itemize} |
| 3663 |
% |
| 3664 |
The following styles are applied in the following order before the cell |
| 3665 |
picture is filled: |
| 3666 |
% |
| 3667 |
\begin{enumerate} |
| 3668 |
\item |/tikz/data visualization/every data set label| with path |
| 3669 |
|/tikz/data visualization| |
| 3670 |
\item |/tikz/data visualization/every label in legend| with path\\ |
| 3671 |
|/tikz/data visualization/legend entry options|. |
| 3672 |
\item The \meta{options}. |
| 3673 |
\item The code in the following key: |
| 3674 |
% |
| 3675 |
\begin{key}{/tikz/data visualization/legend entry options/setup} |
| 3676 |
Some code to be executed at this point. Mostly, it is used to |
| 3677 |
setup attributes for style sheets. |
| 3678 |
\end{key} |
| 3679 |
\item A styling signal is emitted. |
| 3680 |
\item Only for the node: The current value of |node style|. |
| 3681 |
\item Only for the visualizer in legend: The styling that has been |
| 3682 |
accumulated by calls to the following key: |
| 3683 |
% |
| 3684 |
\begin{stylekey}{/tikz/data visualization/legend entry options/visualizer in legend style=\\\marg{options}} |
| 3685 |
Calls to this key accumulate \meta{options} that will be |
| 3686 |
executed with the path prefix |/tikz| at this point. |
| 3687 |
\end{stylekey} |
| 3688 |
\end{enumerate} |
| 3689 |
\end{key} |
| 3690 |
|
| 3691 |
As indicated earlier, the |new legend entry| key is called by the |
| 3692 |
|label in legend=|\meta{options} key internally. In this case, the following |
| 3693 |
extra \meta{extra options} are passed to |new legend entry| key: |
| 3694 |
% |
| 3695 |
\begin{itemize} |
| 3696 |
\item The styling of the visualizer. |
| 3697 |
\item The |/tikz/data visualization/every label in legend| style. |
| 3698 |
\item The |/tikz/every label| style with path |/tikz|. |
| 3699 |
\item Setting |setup| to |/data point/set=|\meta{name of the visualizer}. |
| 3700 |
\item The value of the |label legend options| that are stored in the |
| 3701 |
visualizer. These options can be changed using the following key: |
| 3702 |
% |
| 3703 |
\begin{key}{/tikz/data visualization/visualizer options/label in legend options=\meta{options}} |
| 3704 |
Use this key with a visualizer to configure the label in legend |
| 3705 |
options. Typically, this key is used only internally by a |
| 3706 |
visualizer upon its creating to set the \meta{options} to setup the |
| 3707 |
|visualizer in legend| key. |
| 3708 |
\end{key} |
| 3709 |
\end{itemize} |
| 3710 |
|
| 3711 |
|
| 3712 |
\subsubsection{Reference: Label in Legend Visualizers for Lines and Scatter Plots} |
| 3713 |
|
| 3714 |
Visualizers like |visualize as line| or |visualize as smooth line| use a label |
| 3715 |
in legend visualizer that draws a short line to represent the data set inside |
| 3716 |
the legend. However, this line needs not be a simple straight line, but can be |
| 3717 |
a little curve or a small circle -- indeed, even the default line is not a |
| 3718 |
simple straight line but rather a small zig-zag curve. To configure this line, |
| 3719 |
the two keys are used, although you will only rarely use them directly, but |
| 3720 |
rather use one of the predefined styles mentioned later on. |
| 3721 |
|
| 3722 |
Before we go into the glorious details of all of these keys, let us first have |
| 3723 |
a look at the keys you are most likely to use in practice: The keys for |
| 3724 |
globally reconfiguring the default label in legend visualizers: |
| 3725 |
% |
| 3726 |
\begin{stylekey}{/tikz/data visualization/legend entry options/default label in legend path} |
| 3727 |
This style is set, by default, to |zig zag label in legend line|. It is |
| 3728 |
installed by the styles |straight line|, |smooth line|, and |gap line|, so |
| 3729 |
changing this style will change the appearance of lines in legends. The |
| 3730 |
main other sensible option for this key is |straight label in legend line|. |
| 3731 |
% |
| 3732 |
\begin{codeexample}[width=5cm,preamble={\usetikzlibrary{datavisualization}}] |
| 3733 |
\tikz \datavisualization [ |
| 3734 |
school book axes, visualize as line/.list={a,b}, |
| 3735 |
style sheet=vary dashing, |
| 3736 |
a={label in legend={text=a}}, b={label in legend={text=b}}] |
| 3737 |
data point [x=-1, y=-1, set=a] data point [x=1, y=0, set=a] |
| 3738 |
data point [x=-1, y=1, set=b] data point [x=1, y=0.5, set=b]; |
| 3739 |
\end{codeexample} |
| 3740 |
% |
| 3741 |
\begin{codeexample}[width=5cm,preamble={\usetikzlibrary{datavisualization}}] |
| 3742 |
\tikz \datavisualization [ |
| 3743 |
school book axes, visualize as line/.list={a,b}, |
| 3744 |
legend entry options/default label in legend path/.style= |
| 3745 |
straight label in legend line, |
| 3746 |
style sheet=vary dashing, |
| 3747 |
a={label in legend={text=a}}, b={label in legend={text=b}}] |
| 3748 |
data point [x=-1, y=-1, set=a] data point [x=1, y=0, set=a] |
| 3749 |
data point [x=-1, y=1, set=b] data point [x=1, y=0.5, set=b]; |
| 3750 |
\end{codeexample} |
| 3751 |
% |
| 3752 |
\end{stylekey} |
| 3753 |
% |
| 3754 |
\begin{stylekey}{/tikz/data visualization/legend entry options/default label in legend closed path} |
| 3755 |
This style is executed by |smooth cycle| and |straight cycle|. There are |
| 3756 |
(currently) no other predefined sets of coordinates that can be used |
| 3757 |
instead of the default value |circular label in legend line|. |
| 3758 |
\end{stylekey} |
| 3759 |
|
| 3760 |
\begin{stylekey}{/tikz/data visualization/legend entry options/default label in legend mark} |
| 3761 |
This style is executed by |no lines| and, implicitly, by scatter plots. The |
| 3762 |
default is to use |label in legend line one mark|. Another possible value |
| 3763 |
is |label in legend line three marks|. |
| 3764 |
% |
| 3765 |
\begin{codeexample}[width=5cm,preamble={\usetikzlibrary{datavisualization}}] |
| 3766 |
\tikz \datavisualization [ |
| 3767 |
visualize as scatter/.list={a,b,c}, |
| 3768 |
style sheet=cross marks, |
| 3769 |
legend entry options/default label in legend mark/.style= |
| 3770 |
label in legend three marks, |
| 3771 |
a={label in legend={text=example a}}, |
| 3772 |
b={label in legend={text=example b}}, |
| 3773 |
c={label in legend={text=example c}}]; |
| 3774 |
\end{codeexample} |
| 3775 |
% |
| 3776 |
\end{stylekey} |
| 3777 |
|
| 3778 |
\begin{key}{/tikz/data visualization/legend entry options/label in legend line coordinates=\\\marg{list of coordinates}} |
| 3779 |
This key takes a \meta{list of coordinates}, which are |
| 3780 |
\tikzname-coordinates separated by commas like |(0,0),|\penalty0|(1,1)|. |
| 3781 |
The effect of setting the key is the following: The label in legend |
| 3782 |
visualizer used by, for instance, |visualize as line| will draw a path |
| 3783 |
going through these points. When the line is drawn, the exact same style |
| 3784 |
will be used as was used for the data set. For instance, if the |
| 3785 |
|smooth line| key was used and also the |style=red| key, the line through |
| 3786 |
the \meta{list of coordinates} will also be red and smooth. When the |
| 3787 |
|straight cycle| key was used, the coordinates will also be connected by a |
| 3788 |
cycle, and so on. |
| 3789 |
|
| 3790 |
When the line connecting the \meta{list of coordinates} is drawn, the |
| 3791 |
coordinate system will have been shifted and transformed in such a way that |
| 3792 |
|(0,0)| lies to the left of the text and at half the height of the |
| 3793 |
character ``x''. This means that the right-most-point in the list should |
| 3794 |
usually be |(0,0)| and all other $x$-coordinates should usually be |
| 3795 |
negative. When the |text left| options is used, the coordinate system will |
| 3796 |
have been flipped, so the \meta{list of coordinates} is independent of |
| 3797 |
whether the text is to the right or to the left of the line. |
| 3798 |
|
| 3799 |
Let us now have a look at a first, simple example. We create a legend entry |
| 3800 |
that is just a straight line, so it should start somewhere to the left of |
| 3801 |
the origin at height $0$ and go to the origin: |
| 3802 |
% |
| 3803 |
\begin{codeexample}[width=5cm,preamble={\usetikzlibrary{datavisualization}}] |
| 3804 |
\tikz \datavisualization [ |
| 3805 |
school book axes, visualize as line/.list={a,b}, |
| 3806 |
style sheet=vary dashing, |
| 3807 |
a={label in legend={text=a, |
| 3808 |
label in legend line coordinates={(-1em,0), (0,0)}}}, |
| 3809 |
b={label in legend={text=b, |
| 3810 |
label in legend line coordinates={(-2em,0), (0,0)}}}] |
| 3811 |
data point [x=-1, y=-1, set=a] data point [x=1, y=0, set=a] |
| 3812 |
data point [x=-1, y=1, set=b] data point [x=1, y=0.5, set=b]; |
| 3813 |
\end{codeexample} |
| 3814 |
|
| 3815 |
Now let us make this a bit more fancy and useful by using shifted lines: |
| 3816 |
% |
| 3817 |
\begin{codeexample}[width=5cm,preamble={\usetikzlibrary{datavisualization}}] |
| 3818 |
\tikz \datavisualization [ |
| 3819 |
school book axes, visualize as line/.list={a,b}, |
| 3820 |
legend={up then right}, style sheet=vary dashing, |
| 3821 |
a={label in legend={text=a, |
| 3822 |
label in legend line coordinates={(-2em,-.25ex), (0,0)}}}, |
| 3823 |
b={label in legend={text=b, |
| 3824 |
label in legend line coordinates={(-2em,.25ex), (0,0)}}}] |
| 3825 |
data point [x=-1, y=-1, set=a] data point [x=1, y=0, set=a] |
| 3826 |
data point [x=-1, y=1, set=b] data point [x=1, y=0.5, set=b]; |
| 3827 |
\end{codeexample} |
| 3828 |
|
| 3829 |
In the final example, we use a little ``hat'' to represent lines: |
| 3830 |
% |
| 3831 |
\begin{codeexample}[width=5cm,preamble={\usetikzlibrary{datavisualization}}] |
| 3832 |
\tikz \datavisualization [ |
| 3833 |
school book axes, visualize as line/.list={a,b}, |
| 3834 |
legend={up then right}, style sheet=vary dashing, |
| 3835 |
a={label in legend={text=a, |
| 3836 |
label in legend line coordinates={ |
| 3837 |
(-2em,-.2ex), (-1em,.2ex), (0,-.2ex)}}}, |
| 3838 |
b={label in legend={text=b, |
| 3839 |
label in legend line coordinates={ |
| 3840 |
(-2em,-.2ex), (-1em,.2ex), (0,-.2ex)}}}] |
| 3841 |
data point [x=-1, y=-1, set=a] data point [x=1, y=0, set=a] |
| 3842 |
data point [x=-1, y=1, set=b] data point [x=1, y=0.5, set=b]; |
| 3843 |
\end{codeexample} |
| 3844 |
% |
| 3845 |
\end{key} |
| 3846 |
|
| 3847 |
\begin{key}{/tikz/data visualization/legend entry options/label in legend mark coordinates=\\\marg{list of coordinates}} |
| 3848 |
This key is similar to |label in legend line coordinates|, but now the |
| 3849 |
\meta{list of coordinates} is used as the positions where plot marks are |
| 3850 |
shown. Naturally, plot marks are only shown there if they are also shown by |
| 3851 |
the visualizer in the actual data -- just like the line through the |
| 3852 |
coordinates of the previous key is only shown when there is a line. |
| 3853 |
|
| 3854 |
The \meta{list of coordinates} may be the same as the one used for lines, |
| 3855 |
but usually it is not. In general, it is better to have marks for instance |
| 3856 |
not at the ends of the line. |
| 3857 |
% |
| 3858 |
\begin{codeexample}[width=5cm,preamble={\usetikzlibrary{datavisualization}}] |
| 3859 |
\tikz \datavisualization [ |
| 3860 |
school book axes, visualize as line/.list={a,b}, |
| 3861 |
legend={up then right}, |
| 3862 |
style sheet=vary dashing, |
| 3863 |
style sheet=cross marks, |
| 3864 |
a={label in legend={text=a, |
| 3865 |
label in legend line coordinates={ |
| 3866 |
(-2em,-.2ex), (-1em,.2ex), (0,-.2ex)}, |
| 3867 |
label in legend mark coordinates={ |
| 3868 |
(-1em,.2ex)}}}, |
| 3869 |
b={label in legend={text=b, |
| 3870 |
label in legend line coordinates={ |
| 3871 |
(-2em,-.2ex), (-1em,.2ex), (0,-.2ex)}, |
| 3872 |
label in legend mark coordinates={ |
| 3873 |
(-2em,-.2ex), (0,-.2ex)}}}] |
| 3874 |
data point [x=-1, y=-1, set=a] data point [x=1, y=0, set=a] |
| 3875 |
data point [x=-1, y=1, set=b] data point [x=1, y=0.5, set=b]; |
| 3876 |
\end{codeexample} |
| 3877 |
% |
| 3878 |
\end{key} |
| 3879 |
|
| 3880 |
|
| 3881 |
Naturally, you typically will not give coordinates explicitly for each label, |
| 3882 |
but use one of the following styles: |
| 3883 |
|
| 3884 |
\begin{key}{/tikz/data visualization/legend entry options/straight label in legend line} |
| 3885 |
Just gives a straight line and two plot marks. |
| 3886 |
% |
| 3887 |
\begin{codeexample}[width=5cm,preamble={\usetikzlibrary{datavisualization}}] |
| 3888 |
\tikz \datavisualization [visualize as line, |
| 3889 |
line={style={mark=x}, label in legend={text=example, |
| 3890 |
straight label in legend line}}]; |
| 3891 |
\end{codeexample} |
| 3892 |
% |
| 3893 |
This style might seem like a good idea to use in general, but it does have |
| 3894 |
a huge drawback: Some commonly used plot marks will be impossible to |
| 3895 |
distinguish -- even though there is no problem distinguishing them in a |
| 3896 |
graph. |
| 3897 |
% |
| 3898 |
\begin{codeexample}[width=5cm,preamble={\usetikzlibrary{datavisualization}}] |
| 3899 |
\tikz \datavisualization [visualize as line/.list={a,b,c}, |
| 3900 |
legend entry options/default label in legend path/.style= |
| 3901 |
straight label in legend line, |
| 3902 |
a={style={mark=+}, label in legend={text=bad example a}}, |
| 3903 |
b={style={mark=-}, label in legend={text=bad example b}}, |
| 3904 |
c={style={mark=|}, label in legend={text=bad example c}}]; |
| 3905 |
\end{codeexample} |
| 3906 |
% |
| 3907 |
For this reason, this option is not the default, but rather the next one. |
| 3908 |
\end{key} |
| 3909 |
|
| 3910 |
\begin{key}{/tikz/data visualization/legend entry options/zig zag label in legend line} |
| 3911 |
Uses a small up-down-up line as the label in legend visualizer. The two |
| 3912 |
plot marks are at the extremal points of the line. It works pretty well in |
| 3913 |
almost all situations and is the default. |
| 3914 |
% |
| 3915 |
\begin{codeexample}[width=5cm,preamble={\usetikzlibrary{datavisualization}}] |
| 3916 |
\tikz \datavisualization [ |
| 3917 |
visualize as line=a, |
| 3918 |
visualize as smooth line/.list={b,c}, |
| 3919 |
a={style={mark=+}, label in legend={text=better example a}}, |
| 3920 |
b={style={mark=-}, label in legend={text=better example b}}, |
| 3921 |
c={style={mark=|}, label in legend={text=better example c}}]; |
| 3922 |
\end{codeexample} |
| 3923 |
% |
| 3924 |
Even though the above example shows that the marks are easier to |
| 3925 |
distinguish than with a straight line, the chosen marks are still not |
| 3926 |
optimal. This is the reason that the |cross marks| style uses different |
| 3927 |
crosses: |
| 3928 |
% |
| 3929 |
\begin{codeexample}[width=5cm,preamble={\usetikzlibrary{datavisualization}}] |
| 3930 |
\tikz \datavisualization [ |
| 3931 |
visualize as line/.list={a,b}, |
| 3932 |
visualize as smooth line=c, |
| 3933 |
style sheet=cross marks, |
| 3934 |
a={label in legend={text=good example a}}, |
| 3935 |
b={label in legend={text=good example b}}, |
| 3936 |
c={gap line, label in legend={text=good example c}}]; |
| 3937 |
\end{codeexample} |
| 3938 |
% |
| 3939 |
\end{key} |
| 3940 |
|
| 3941 |
\begin{key}{/tikz/data visualization/legend entry options/circular label in legend line} |
| 3942 |
This style is especially tailored to represent lines that are closed. It is |
| 3943 |
automatically selected for instance by the |polygon| or the |smooth cycle| |
| 3944 |
styles. |
| 3945 |
% |
| 3946 |
\begin{codeexample}[ |
| 3947 |
width=7cm, |
| 3948 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 3949 |
] |
| 3950 |
\tikz \datavisualization [ |
| 3951 |
scientific axes={clean}, all axes={length=3cm}, |
| 3952 |
visualize as line/.list={a,b,c}, |
| 3953 |
a={polygon}, b={smooth cycle}, |
| 3954 |
style sheet=cross marks, |
| 3955 |
a={label in legend={text=polygon}}, |
| 3956 |
b={label in legend={text=circle}}, |
| 3957 |
c={label in legend={text=line}}] |
| 3958 |
data [format=function, set=a] { |
| 3959 |
var t : {0,72,...,359}; |
| 3960 |
func x = cos(\value t); |
| 3961 |
func y = sin(\value t); |
| 3962 |
} |
| 3963 |
data [format=function, set=b] { |
| 3964 |
var t : [0:2*pi]; |
| 3965 |
func x = .8*cos(\value t r); |
| 3966 |
func y = .8*sin(\value t r); |
| 3967 |
} |
| 3968 |
data point [x=-1, y=0.5, set=c] |
| 3969 |
data point [x=1, y=0.25, set=c]; |
| 3970 |
\end{codeexample} |
| 3971 |
% |
| 3972 |
\end{key} |
| 3973 |
|
| 3974 |
\begin{key}{/tikz/data visualization/legend entry options/gap circular label in legend line} |
| 3975 |
This style is especially tailored to for the |gap cycle| style and |
| 3976 |
automatically selected by it: |
| 3977 |
% |
| 3978 |
\begin{codeexample}[ |
| 3979 |
width=7cm, |
| 3980 |
preamble={\usetikzlibrary{datavisualization.formats.functions}}, |
| 3981 |
] |
| 3982 |
\tikz \datavisualization [ |
| 3983 |
scientific axes={clean}, all axes={length=3cm}, |
| 3984 |
visualize as line/.list={a,b,c}, |
| 3985 |
a={gap cycle}, b={smooth cycle}, c={gap line}, |
| 3986 |
a={style={mark=*, mark size=0.5pt}, |
| 3987 |
label in legend={text=polygon}}, |
| 3988 |
b={label in legend={text=circle}}, |
| 3989 |
c={style={mark=*, mark size=0.5pt, mark options=red}, |
| 3990 |
label in legend={text=line}}] |
| 3991 |
data [format=function, set=a] { |
| 3992 |
var t : {0,72,...,359}; |
| 3993 |
func x = cos(\value t); |
| 3994 |
func y = sin(\value t); |
| 3995 |
} |
| 3996 |
data [format=function, set=b] { |
| 3997 |
var t : [0:352]; |
| 3998 |
func x = .8*cos(\value t); |
| 3999 |
func y = .8*sin(\value t); |
| 4000 |
} |
| 4001 |
data point [x=-1, y=0.5, set=c] |
| 4002 |
data point [x=1, y=0.25, set=c]; |
| 4003 |
\end{codeexample} |
| 4004 |
% |
| 4005 |
\end{key} |
| 4006 |
|
| 4007 |
\begin{key}{/tikz/data visualization/legend entry options/label in legend one mark} |
| 4008 |
To be used with scatter plots, since no line is drawn. Just displays a |
| 4009 |
single mark (this is the default with a scatter plot or when the |no line| |
| 4010 |
is selected. |
| 4011 |
% |
| 4012 |
\begin{codeexample}[width=5cm,preamble={\usetikzlibrary{datavisualization}}] |
| 4013 |
\tikz \datavisualization [visualize as scatter/.list={a,b,c}, |
| 4014 |
style sheet=cross marks, |
| 4015 |
a={label in legend={text=example a}}, |
| 4016 |
b={label in legend={text=example b}}, |
| 4017 |
c={label in legend={text=example c}}]; |
| 4018 |
\end{codeexample} |
| 4019 |
% |
| 4020 |
\end{key} |
| 4021 |
|
| 4022 |
\begin{key}{/tikz/data visualization/legend entry options/label in legend three marks} |
| 4023 |
An alternative to the previous style, where several marks are shown. |
| 4024 |
% |
| 4025 |
\begin{codeexample}[width=5cm,preamble={\usetikzlibrary{datavisualization}}] |
| 4026 |
\tikz \datavisualization [visualize as scatter/.list={a,b,c}, |
| 4027 |
style sheet=cross marks, |
| 4028 |
a={label in legend={text=example a, label in legend three marks}}, |
| 4029 |
b={label in legend={text=example b, label in legend three marks}}, |
| 4030 |
c={label in legend={text=example c, label in legend three marks}}]; |
| 4031 |
\end{codeexample} |
| 4032 |
% |
| 4033 |
\end{key} |