Re: [igraph] Parse error in Pajek format

[Gábor Csárdi]

Fabio,

I am sorry, I overlooked your question, just noticed it now, that I'm
fixing the Pajek reader for this issue.

Some vertex and edge attributes are saved in Pajek files, not all of
them, because Pajek does not support arbitrary attributes, AFAIK. See
the documentation at ?read.graph for the list of attributes that are
saved.

Thanks again for your report, I've now fixed this issue in the igraph
development tree.

Best,
Gabor

On Wed, Nov 17, 2010 at 6:47 PM, Fabio Daolio <address@hidden> wrote:
> Dear Gabor,
>
> thank for the quick reply! I think you're right: checking in the manual
> http://vlado.fmf.uni-lj.si/pub/networks/pajek/doc/pajekman.pdf
> there's no sign of a number after Arcs* (pp. 72-75).
>
> That is the only documentation I found. Actually I'm not on Windows
> and I don't use Pajek: I was just trying to use the InfoMap algorithm
> http://www.tp.umu.se/~rosvall/code.html#map
> which requires networks to be in Pajek format.
> I found that network among the examples and I thought it was legal
> because both Gephy and Visone were able to open it, no problem.
>
> To the contrary, now I see that the InfoMap program is not happy
> with the .net files that are produced by write.graph(), so maybe the
> issue was due to their implementation of the Pajek format...
>
> Anyway, could it be possible to have write.graph() saving the label
> and attributes of vertices? I'm asking because in my version it only
> saves the line: *Vertices <n>
> and then starts with the edge list, completely skipping the vertex lists.
>
> Always thankful,
> all the best
> --
> Fabio
>
> On 17 nov. 2010, at 17:19, Gábor Csárdi wrote:
>
>> Dear Fabio,
>>
>> the problem is that there is no definition at all for the Pajek file
>> format. (As far as I know.)
>> Plus there is no source code for Pajek either, that we could use to
>> see what formats Pajek accepts.
>>
>> So far I have never seen Pajek files with a number after *Arcs, but if
>> you say this is legal, then we can easily support it, of course. I'll
>> add this to the development version, in a minute.
>>
>> Thanks for your report! Best,
>> Gabor
>>
>> On Wed, Nov 17, 2010 at 5:13 PM, Fabio Daolio <address@hidden> wrote:
>>> Dear all,
>>>
>>> I'm using igraph 0.5.5 together with R 2.12.0
>>> I noticed that read.graph() on a .net file fails with a parse error
>>> whenever the file contains a line like
>>>
>>> *Arc <x>
>>>
>>> where <x> should be the number of edges, I guess.
>>> Removing that number seems to solve the problem, even if the
>>> original .net was a legal Pajek file... (attached)
>>> Could it be an issue with the parsing in igraph? I know you
>>> support just a subset of that file format, anyway I wanted to
>>> tell everybody in case someone else might find it useful.
>>>
>>> Thanks once again for your good work!
>>> --
>>> Fabio Daolio
>>>
>>>
>>> _______________________________________________
>>> igraph-help mailing list
>>> address@hidden
>>> http://lists.nongnu.org/mailman/listinfo/igraph-help
>>>
>>>
>>
>>
>>
>> --
>> Gabor Csardi <address@hidden>     UNIL DGM
>>
>> _______________________________________________
>> igraph-help mailing list
>> address@hidden
>> http://lists.nongnu.org/mailman/listinfo/igraph-help
>
>
>
> _______________________________________________
> igraph-help mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/igraph-help
>



-- 
Gabor Csardi <address@hidden>     UNIL DGM