March 2014 - graph-tool - archives.skewed.de

New to graphtool , coming from networkx. Any support for json?
by phyo.arkarlwin＠gmail.com 21 Mar '14

21 Mar '14

Hello Graph-Tool I had experience with networkx , but due to performance requirements i am thinking about trying graph-tool. My main goal it to generate network graph for email communications , but i need to make it available online as interactive networks so i want to know if json serialization available? I will use sigma.js http://sigmajs.org/ to render network graphs like this : http://exploringdata.github.io/info/npm-packages-dependencies/ I would be very grateful if i can have pointers for that. Thanks Phyo.

3 6

is getting the adjacency matrix costly?
by Hang Mang 20 Mar '14

20 Mar '14

Is it costly to repeatedly access the adjacency matrix with graph_tool.spectral.adjacency() ? i.e. does it get computed every time I make a call to it? If so, is it really costly to do?

2 1

checking if an edge exist?
by Hang Mang 20 Mar '14

20 Mar '14

How can I check if a edge exists between a source and target or not?!

2 1

graph_union, unexpected behavior
by Stefan Klingelschmitt 20 Mar '14

20 Mar '14

Hey, I'm struggeling with the grap_union() function. It seems like the properties are only correctly passed to the union graph, if include=True. Is this how it supposed to work? Here a minimal example: import graph_tool.all as gt import graph_tool.topology as gt_topo from gi.repository import Gtk, Gdk, GdkPixbuf, GObject G = gt.Graph(directed=False) G_sub = gt.Graph(directed=False) v1 = G.add_vertex() v2 = G.add_vertex() v3 = G.add_vertex() v4 = G_sub.add_vertex() v5 = G_sub.add_vertex() e1 = G.add_edge(v2, v1) e2 = G.add_edge(v2, v3) e3 = G_sub.add_edge(v5, v4) G.vertex_properties['type'] = G.new_vertex_property("string") G_sub.vertex_properties['type'] = G_sub.new_vertex_property("string") G.vertex_properties['type'][v1] = 'Car' G.vertex_properties['type'][v2] = 'Crossing' G.vertex_properties['type'][v3] = 'Car' G_sub.vertex_properties['type'][v4] = 'Pedestrian' G_sub.vertex_properties['type'][v5] = 'Sidewalk' #ToDo: Function bug, why is include=True needed? G_union, prop = gt.graph_union(G, G_sub, props=[(G.vertex_properties['type'], \ G_sub.vertex_properties['type'])], include=True) pos = gt.sfdp_layout(G) gt.graph_draw(G_union, pos, vertex_text=G_union.vertex_properties['type']) Another point that I find a bit not so intuitive is that you have to specify every single property of the two graphs you would like to have in the new graph, wouldn't it be much more intuitive to copy all the properties into the new graph unless not told otherwise? P.S: This will probably not my last question as I'm planning to heavily use graph-tool for my research....

2 3

Error: Locale not supported by C library
by Hang Mang 19 Mar '14

19 Mar '14

I'm using Eclipse and whenever I run the following import: "import graph_tool.draw as gt_draw" I get the following error in the console: "(process:4178): Gtk-WARNING **: Locale not supported by C library. Using the fallback 'C' locale." What's wrong please?

2 1

Drawing in IPython Notebook inline (web interface)
by Joel Moberg 14 Mar '14

14 Mar '14

I don't know if this feature exists already, I haven't found it atleast. In IPython notebook it would be nice to draw a graph when calling `graph_draw(`. I found how igraph did some of this here: https://github.com/igraph/igraph/pull/7/files.

2 1

Keyserver OK?
by Tim Day 14 Mar '14

14 Mar '14

I'm having problems accessing the keyfile for the Debian/Ubuntu packages via the link on the download page: http://pgp.skewed.de:11371/pks/lookup?op=get&search=0x612DEFB798507F25 gets me an "unable to connect" page in Firefox. I can ping pgp.skewed.de fine. Just me? (I'm travelling at the moment, I don't think the internet here is blocking any ports but not sure how to tell for sure). Can the keyfile be obtained through any other public keyservers? (I did try searching the PGP Global Directory for 98507F25 but no joy). Thanks for any info/help Tim

2 3

behaviour of blockmodel, Potts with disconnected graph
by Alan Williams 13 Mar '14

13 Mar '14

Hi Tiago, how do the blockmodel and modularity community functions behave with a graph with one giant component and a few relatively tiny disconnected components? Do I need to create a new graph of just the dominant component and then run the community functions on that, or will the functions lump the disconnected vertices into a block/community of the dominant component with no noticeable effect on the algorithm outcome? Thanks! Alan

2 1

wrong label_largest_component
by Николай Кинаш 13 Mar '14

13 Mar '14

Hello. import sys, os from pylab import * from numpy.random import * from graph_tool.all import * g = Graph(directed=False); N = 10; v = g.add_vertex(); vlist = [v]; for i in range(1, N): v = g.add_vertex() c = randint(0, len(vlist)) target = vlist[c] g.add_edge(v, target) vlist.append(v) if randint(0, 2) == 1: rem = randint(0, len(vlist)) g.remove_vertex( vlist[rem] ); graph_tool.stats.remove_self_loops(g); pos = sfdp_layout(g, cooling_step=0.99); graph_draw(g, pos, output_size=(1000, 1000),vertex_size=10, edge_pen_width=2.2, output="graph" + str(i) + ".png") l = graph_tool.topology.label_largest_component(g); print(l.a); u = graph_tool.topology.GraphView(g, vfilt=l); print(u.num_vertices()); =============out======= python test_graph.py [0 1 1 0 0] 2 but in image i see 4 vertex in largest connected component. I try it over and over, and always see that 1 or 2 vertex missed.

2 2

Adjancymatrix graphcreation
by v k 13 Mar '14

13 Mar '14

Hey folks, Is there an efficient way to create a graph out of a sparse adjancy matrix? Otherwise my idea is rather to use np.nonzero and iterate over it all, but it didnt naive to me. best, vsilv

2 1