Better figure for illustrating matching algs.

[Alpar Juttner]

The doc of module "Matching Algorithms" presents a bipartite graph as and illustration for the result of these algorithm. This falsely suggests that these algs work on this kind of graphs. In fact, the opposite is true, the bipartite graphs - therefore the bipartite matching algorithms - are yet to be ported from the 0.x series.

Thus, I would prefer an example on a general graph here.

[Peter Kovacs]

An alternative solution: remove this image in the release branch 1.2, but keep it in the main branch, since it will be good once we have bipartite graphs and bipartite matchings (probably in release 1.3).

[Balazs Dezso]

I have uploaded two files, are they proper for replace the current figure?

The graph is generated with lgf_gen, and the matching is executed. The source of the drawing can be found:

​http://lime.cs.elte.hu/~deba/open/hgwebdir.cgi/deba_lemon_loc/file/d38d9083cc28/draw_max_matching.cc

[Alpar Juttner]

Replying to deba:

​http://lime.cs.elte.hu/~deba/open/hgwebdir.cgi/deba_lemon_loc/file/d38d9083cc28/draw_max_matching.cc

Thanks. Based on your code I made an other figure​. Less nodes, with some adjustments by hand. More articulated nodes, wider edges, red matching.

How do you like it?

[Peter Kovacs]

Replying to alpar:

Thanks. Based on your code I made an other figure​. Less nodes, with some adjustments by hand. More articulated nodes, wider edges, red matching.

How do you like it?

I like both images, but Alpar's version seems better for the doc.

[Alpar Juttner]

Replying to kpeter:

I like both images, but Alpar's version seems better for the doc.

The figure has been changed in the repo, see [d48d79b11f5b].