1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
|
<!--<!DOCTYPE chapter SYSTEM "http://www.oasis-open.org/docbook/xml/4.4/docbookx.dtd">-->
<chapter id="tutorial_beginner"><title>Tutorial</title>
<sect1 id="first-step"><title>First Steps</title>
<para> The goal of this part is to create a graph, save it, and apply the most common analysis operations.</para>
<sect2 id="first-dis"><title>Creating a graph</title>
There are several ways to create a graph: through a generation alorithm (see below), by manually building it element by element (see below),
or through the wizard "Import CSV Data" (see <ulink url="http://tulip.labri.fr/TulipDrupal/?q=node/1411">this video tutorial</ulink>).
<sect3 id="import"><title>Generate a graph</title>
<para>
<itemizedlist>
<listitem><para>Click menu <code>File → Import → Graph</code>.</para>
</listitem>
<listitem><para>Select a generation method, for instance <code>Grid approximation</code> and click OK.</para></listitem>
</itemizedlist>
A graph with nodes and edges is now displayed.
With the mouse wheel, you can zoom in an out.
Click and drag to move the graph.
Shift-click rotates the graph around x and y axis, Ctrl-click rotates around the z axis.
To reset the view, right-click and select <code>View → Center view</code>.
</para>
</sect3>
<sect3 id="create"><title>Build a graph</title>
<para>
<itemizedlist>
<listitem> <para>Click Menu <code>File → New</code>.</para> </listitem>
<listitem>
<para> Select the <inlinegraphic fileref="images/i_addedge.png"/> tool.
Click on a few different places in the "Node Link Diagram View" subwindow: each click adds a node.
Now, add a few links: for each link, click on the source node, then on the target node.</para>
</listitem>
<listitem>
Click <code>View Editor</code> (bottom left corner) → <code>Rendering parameters → arrows</code> to display or hide them.
Click <code>Layer manager</code> and uncheck <code>Edges</code> then click <code>Apply</code> to hide the edges.
The first column (visibility) determines whether this element is visible,
the second column determines stencil: when a layer has the stencil, it is displayed on top of layers who do not.
</listitem>
<listitem>
<para>To delete an element, select the <inlinegraphic fileref="images/i_del.png"/> tool and click the element.</para>
</listitem>
</itemizedlist>
</para>
</sect3>
</sect2>
<sect2 id="save"><title>Saving</title>
<para>To save your work in Tulip's own file format (.tlp), click menu <code>File → Save</code> or <code>File → Save as</code>.
To save it in another file format (only GML and TLP are supported without any plugin installed),
click menu <code>File → Export</code>.</para>
</sect2>
<sect2 id="algos"><title>Algorithms</title>
<para>
Generate a graph: click Menu <code>File → Import → Graph → Random General Graph</code>, type in "50" in the nodes field and "50" edges.
Since node positions are random, the layout is tangled:
<graphic fileref="images/tutorial_beginner_random-graph.png"/>
To improve it, let's apply a <emphasis>Layout algorithm</emphasis>: click <code>Algorithm → Layout → Force directed → GEM</code>.
It computes a good position for each node:
<graphic fileref="images/tutorial_beginner_random-graph_laid-out.png"/>
</para>
<para>
Some <code>layout</code> algorithms can only be applied to graphs belonging to a specific graph class.
For instance, the <code>Bubble Tree</code> algorithm can only be applied on a tree.
There are algortihms for computing Layout, Measures, Selection, Clustering (in <code>General</code>) and more.
</para>
</sect2>
<sect2><title>Properties</title>
Properties are an important concept in Tulip, they are used to store information about each node and edge. To see them,
click <code>Graph Editor</code> (bottom left corner) → <code>Properties</code> (tab on the left),
and in the bottom part click <code>viewColor</code>.
This property stores, for each node, the color used to display it. In the upper part,
double-click a red cell and choose a blueish color, click OK,
and press <keycap>Enter</keycap> to update the graph display on the right:
<graphic fileref="images/tutorial_beginner_viewColor.png"/>
The corresponding node (in the picture, the node with id 2) is displayed in blue (on the drawing, it is in the bottom right corner).
<para>To compute more meaningful colors, click menu <code>Algorithm → Measure → Graph → Betweeness centrality</code>,
this computes a centrality for each node. The graph is now displayed with colors depending on this metric
(nodes at the center have a high centrality and are thus blue, node at the border have a low centrality and are thus red):
<graphic fileref="images/tutorial_beginner_centrality.png"/>
The result of this computation has been stored in the viewMetric property:
<graphic fileref="images/tutorial_beginner_viewMetric.png"/>
But if we run another algorithm, it will overwrite the viewMetric property. So, let's copy the result into another property:
click <code>viewMetric</code> then <code>copy</code> and type in "centrality" (as a new property).
</para>
<para>
Now, let's add labels. In the bottom part (the list of properties), click <code>viewLabel</code>
(not "<code>viewLabel...</code>"),
then double-click the cell under the mouse in the picture below, and type in "Hello World !".
Click OK, then press <keycap>Enter</keycap> to update the display. The node with id 2 has a label:
<graphic fileref="images/tutorial_beginner_viewLabel.png"/>
To get more meaningful labels, let's apply an algorithm.
Click <code>Algorithm → Measure → Misc → Id</code>. The result of this metric (here, the identifier of each node)
has been stored in the viewMetric property (overwriting the previously computed centrality).
Also, new colors have been computed to represent this property, but do not pay attention to them for now:
<graphic fileref="images/tutorial_beginner_id.png"/>
To display this viewMetric property as labels, click <code>viewMetric</code> then <code>To labels</code>.
(One can see that the colors represent the id: small id have a reddish color, while high id have a blueish color.)
To restore the colors representing centralilty, click
<code>Algorithm → Color → Color mapping</code> and select "centrality" under "linear/uniform property". Click <code>OK</code>.
The final picture should look like this (since the graph is random, the exact structure can be different):
<graphic fileref="images/tutorial_beginner_id_centrality.png"/>
</para>
</sect2>
</sect1>
</chapter>
|
