[Debian-med-packaging] Packaging R igraph for Debian

[Andreas Tille]

Hi Gábor,

On Fri, Sep 26, 2014 at 09:51:17AM -0400, Gábor Csárdi wrote:
> the first one is an issue that has come up a couple of times so far
> with ARPACK, but I don't know how to fix it in this case, I can't seem
> to be able to reproduce it.

Hmmmm, may be I might exclude this test for the moment.
 
> The second one I can reproduce, and it will be fixed in the next
> version of igraph.

Fine.  Do you have a rough estimation when this will be available?  I do
not want to put some pressure on you - just want to know whether I
should wait or for the moment document the problem and skip the test.

Thanks for your quick feedback

        Andreas.

> Best,
> Gabor
> 
> On Thu, Sep 25, 2014 at 4:35 PM, Andreas Tille <andreas at an3as.eu> wrote:
> > Hi Gabor,
> >
> > I intend to package R igraph for Debian[1] since it is needed by one of
> > the target packages of the Debian Med project.  When I run the unit
> > tests via
> >
> > LC_ALL=C R --no-save <<EOT
> > library(igraph)
> > igraphtest()
> > EOT
> >
> >
> > I get
> >
> >
> > walktrap.community : ........
> > watts.strogatz.game : ..................................................
> >
> >
> > 1. Error: evcent works -----------------------------------------------------
> > At arpack.c:944 : ARPACK error, No shifts could be applied during a cycle of the Implicitly restarted Arnoldi iteration. One possibility is to increase the size of NCV relative to NEV
> > 1: evcent(G) at test_evcent.R:42
> > 2: .Call("R_igraph_eigenvector_centrality", graph, directed, scale, weights, options,
> >        PACKAGE = "igraph")
> > 3: .handleSimpleError(function (e)
> >    {
> >        e$calls <- head(sys.calls()[-seq_len(frame + 7)], -2)
> >        signalCondition(e)
> >    }, "At arpack.c:944 : ARPACK error, No shifts could be applied during a cycle of the Implicitly restarted Arnoldi iteration. One possibility is to increase the size of NCV relative to NEV",
> >        quote(.Call("R_igraph_eigenvector_centrality", graph, directed, scale, weights,
> >            options, PACKAGE = "igraph")))
> >
> > 2. Error: leading.eigenvector.community works ------------------------------
> > (list) Objekt kann nicht nach 'double' umgewandelt werden
> > 1: leading.eigenvector.community(g, callback = f) at test_leading.eigenvector.community.R:26
> > 2: .Call("R_igraph_community_leading_eigenvector", graph, steps, weights, options, start,
> >        callback, extra, env, environment(igraph.i.levc.arp), PACKAGE = "igraph")
> > 3: .handleSimpleError(function (e)
> >    {
> >        e$calls <- head(sys.calls()[-seq_len(frame + 7)], -2)
> >        signalCondition(e)
> >    }, "(list) Objekt kann nicht nach 'double' umgewandelt werden", quote(.Call("R_igraph_community_leading_eigenvector",
> >        graph, steps, weights, options, start, callback, extra, env, environment(igraph.i.levc.arp),
> >        PACKAGE = "igraph")))
> >
> >
> > Any hint how to run the test suite successfully without these errors?
> >
> > Kind regards
> >
> >       Andreas.
> >
> >
> >
> > [1] http://anonscm.debian.org/viewvc/debian-med/trunk/packages/R/r-cran-igraph/trunk/
> >
> > --
> > http://fam-tille.de
> 

-- 
http://fam-tille.de