Segmentation fault when calculating eigenvector centrality

[Amy Yu]

Python 2.6.5 (r265:79063, Apr 16 2010, 13:57:41)
In [1]: import igraph
In [2]: g = igraph.Graph([(0,1), (0,2), (2,3), (3,4), (4,2), (2,5), (5,0), (6,3), (5,6)])
In [3]: g.eigenvector_centrality()
Segmentation fault (core dumped)

I get a segmentation fault error whenever I attempt to calculate eigenvector centrality using igraph. I am using version 0.5.4.

[Tamás Nepusz (ntamas)]

What OS are you using and how did you install igraph? The reason why I am asking is because it works for me with igraph 0.5.4 (compiled fresh from the source code):

tamas@keszeg:~/bzr/igraph/0.5-main/build $ igraph
No configuration file, using defaults
igraph 0.5.4 running inside Python 2.6.6 (r266:84292, Sep 15 2010, 16:22:56)
Type "copyright", "credits" or "license" for more information.

IPython 0.10 -- An enhanced Interactive Python.
? -> Introduction and overview of IPython's features.
%quickref -> Quick reference.
help -> Python's own help system.
object? -> Details about 'object'. ?object also works, ?? prints more.

In [1]: g=Graph([(0,1), (0,2), (2,3), (3,4), (4,2), (2,5), (5,0), (6,3), (5,6)])
In [2]: g.eigenvector_centrality()
Out[2]:
[0.71119954501079996,
 0.24860112203602616,
 1.0,
 0.75139887796397364,
 0.6122047310756773,
 0.78600269906125164,
 0.53740157702522562]

[Amy Yu (amyzhao-yu)]

Hi Tamás,

Thanks for the fast response! I am using Ubuntu 10.04 LTS. I installed
igraph by doing:

sudo add-apt-repository ppa:igraph/ppa

sudo aptitude update
sudo aptitude install python-igraph

I just tried reinstalling using the above commands but I am still
getting the segmentation fault. Let me know if you have any
suggestions!

Thanks,

Amy

<http://igraph.wikidot.com/installing-python-igraph-on-linux>

On Tue, Mar 22, 2011 at 5:50 AM, Tamás Nepusz <email address hidden> wrote:

> What OS are you using and how did you install igraph? The reason why I
> am asking is because it works for me with igraph 0.5.4 (compiled fresh
> from the source code):
>
> tamas@keszeg:~/bzr/igraph/0.5-main/build $ igraph
> No configuration file, using defaults
> igraph 0.5.4 running inside Python 2.6.6 (r266:84292, Sep 15 2010,
> 16:22:56)
> Type "copyright", "credits" or "license" for more information.
>
> IPython 0.10 -- An enhanced Interactive Python.
> ? -> Introduction and overview of IPython's features.
> %quickref -> Quick reference.
> help -> Python's own help system.
> object? -> Details about 'object'. ?object also works, ?? prints more.
>
> In [1]: g=Graph([(0,1), (0,2), (2,3), (3,4), (4,2), (2,5), (5,0), (6,3),
> (5,6)])
> In [2]: g.eigenvector_centrality()
> Out[2]:
> [0.71119954501079996,
> 0.24860112203602616,
> 1.0,
> 0.75139887796397364,
> 0.6122047310756773,
> 0.78600269906125164,
> 0.53740157702522562]
>
> --
> You received this bug notification because you are a direct subscriber
> of the bug.
> https://bugs.launchpad.net/bugs/739954
>
> Title:
> Segmentation fault when calculating eigenvector centrality
>
> Status in The igraph library:
> New
>
> Bug description:
> Python 2.6.5 (r265:79063, Apr 16 2010, 13:57:41)
> In [1]: import igraph
> In [2]: g = igraph.Graph([(0,1), (0,2), (2,3), (3,4), (4,2), (2,5), (5,0),
> (6,3), (5,6)])
> In [3]: g.eigenvector_centrality()
> Segmentation fault (core dumped)
>
> I get a segmentation fault error whenever I attempt to calculate
> eigenvector centrality using igraph. I am using version 0.5.4.
>
> To unsubscribe from this bug, go to:
> https://bugs.launchpad.net/igraph/+bug/739954/+subscribe
>

--
-Amy

[Tamás Nepusz (ntamas)]

Is it a 64-bit or a 32-bit system?

[Tamás Nepusz (ntamas)]

Seems to happen on 64-bit systems only; I'll try to push out a new version of the package that should work. It will take a couple of hours, though.

[Tamás Nepusz (ntamas)]

The fixed package is currently being re-built by Launchpad. If you wish, I can send you a package built on my own machine so you can try it. Just drop me a mail.

[Amy Yu (amyzhao-yu)]

Sure! Do I install it the same way?

Via mobile

On Mar 22, 2011, at 12:20 PM, "Tamás Nepusz" <email address hidden> wrote:

> The fixed package is currently being re-built by Launchpad. If you wish,
> I can send you a package built on my own machine so you can try it. Just
> drop me a mail.
>
> --
> You received this bug notification because you are a direct subscriber
> of the bug.
> https://bugs.launchpad.net/bugs/739954
>
> Title:
> Segmentation fault when calculating eigenvector centrality
>
> Status in The igraph library:
> Confirmed
>
> Bug description:
> Python 2.6.5 (r265:79063, Apr 16 2010, 13:57:41)
> In [1]: import igraph
> In [2]: g = igraph.Graph([(0,1), (0,2), (2,3), (3,4), (4,2), (2,5), (5,0), (6,3), (5,6)])
> In [3]: g.eigenvector_centrality()
> Segmentation fault (core dumped)
>
> I get a segmentation fault error whenever I attempt to calculate
> eigenvector centrality using igraph. I am using version 0.5.4.
>
> To unsubscribe from this bug, go to:
> https://bugs.launchpad.net/igraph/+bug/739954/+subscribe

[Tamás Nepusz (ntamas)]

For the record: Amy wrote that the problem is solved with the new Debian package. It was an issue with the usage of an external ARPACK implementation on a 64-bit machine - this will be fixed permanently in the next release. The 0.6-arpack branch was already merged into trunk.

[Jordi Torrents (jtorrents)]

Hi Tamas,

I've just installed the igraph python packages from PPA Launchpad (build 5 weeks ago) and it seems that the bug is still there:

jtorrents@jupiter:~$ igraph
No configuration file, using defaults
igraph 0.5.4 running inside Python 2.6.6 (r266:84292, Dec 26 2010, 22:31:48)
Type "copyright", "credits" or "license" for more information.

IPython 0.10 -- An enhanced Interactive Python.
? -> Introduction and overview of IPython's features.
%quickref -> Quick reference.
help -> Python's own help system.
object? -> Details about 'object'. ?object also works, ?? prints more.

In [1]: g=Graph([(0,1), (0,2), (2,3), (3,4), (4,2), (2,5), (5,0), (6,3), (5,6)])
In [2]: g.eigenvector_centrality()
Segmentation fault
jtorrents@jupiter:~$

I'm using Debian Squeeze 64bits (amd64) and I have used the maverik target on PPA. In a system with the same configuration but with 32 bits (ie Debian stable and igraph installed from maverik target on PPA) seems to work fine.

Thanks

[Jordi Torrents (jtorrents)]

Hi again,

It seems that the problem is that the build with Tamas' fix (libigraph0 0.5.4-4) is only available for Lucid and not for Maverick. Installing libigraph0 0.5.4-4 solved the segfault in my system (Debian Squeeze amd64).

Salut!

[Tamás Nepusz (ntamas)]

Yes, sorry, I forgot to copy the new packages from the lucid repo to maverick & natty. This has to be done manually after the build finishes, and sometimes I forget this. Thanks for the heads up.

[Gábor Csárdi (gabor.csardi)]

The development of igraph has moved to github, so please do not comment on this bug here. You are of course welcome to comment on github, here:
https://github.com/igraph/igraph/issues/87