Static sparse graphs for fast low-level computations

[nathanncohen]

Helloooooo !!

This extensively documented module implements a very basic data structure for graphs that is helpful for EFFICIENT implementations. It was actually used by Sage already in sage.graphs.distances_all_pairs, but it is better to have a proper documentation for such things.

And of course, this does not solve the current lack of a Python-level static graph class, that would handle loops/multiedges and labels... It will come, though :-)

Nathann

Depends on #12235

CC: @dcoudert

Component: graph theory

Keywords: Cernay2012

Author: Nathann Cohen

Reviewer: David Coudert

Merged: sage-5.0.beta5

Issue created by migration from https://trac.sagemath.org/ticket/12306

[dcoudert]

comment:2

I'm unable to install the patch on sage-5.0.beta1.

File digraph.py.rej:

--- digraph.py
+++ digraph.py
@@ -2565,18 +2574,24 @@
 
         TESTS:
 
-        Checking against NetworkX::
-
+        Checking against NetworkX, and another of Sage's implementations::
+
+            sage: from sage.graphs.base.static_sparse_graph import strongly_connected_components
             sage: import networkx
             sage: for i in range(100):                                     # long
             ...        g = digraphs.RandomDirectedGNP(100,.05)             # long
             ...        h = g.networkx_graph()                              # long
             ...        scc1 = g.strongly_connected_components()            # long
             ...        scc2 = networkx.strongly_connected_components(h)    # long
+            ...        scc3 = strongly_connected_components(g)             # long
             ...        s1 = Set(map(Set,scc1))                             # long
             ...        s2 = Set(map(Set,scc2))                             # long
+            ...        s3 = Set(map(Set,scc3))                             # long
             ...        if s1 != s2:                                        # long
             ...            print "Ooch !"                                  # long
+            ...        if s1 != s3:                                        # long
+            ...            print "Oooooch !"                               # long
+
         """
 
         try:

[nathanncohen]

comment:3

Oopps.. This patch needs #12235 to be applied first, though it is now merged with beta2 :-)

Nathann

[dcoudert]

comment:4

I can now install the patch with sage-5.0.beta2.

No compilation error. Generation of the documentation is OK.

However, can you improve in section What does it take as input? the n^2sizeof(...) and the nsizeof(...). For instance you can add a \cdot between n and sizeof(...).

For functions returning a dictionnary of dictionnary, you should add an extra warning recalling that such structures are huge. I'm not sure most of us can use this for graphs with 10.000 nodes.

Best,

D.

[dcoudert]

Reviewer: David Coudert

[nathanncohen]

comment:5

Here it is !!! :-)

I also removed some trailing whitespaces, as I learned here that they were evil :-P

Nathann

[nathanncohen]

Changed keywords from none to Cernay2012

[nathanncohen]

Attachment: trac_12306_doc.patch.gz

[dcoudert]

comment:6

I can install both patch correctly and the documentation is now well presented and with enough details.

Good work !

D.

PS: what's the relation between this patch and the Cernay 2012 Music Festival ? ;-)

[nathanncohen]

comment:7

PS: what's the relation between this patch and the Cernay 2012 Music Festival ? ;-)

Some snow, and around 10 people in a tower coding Sage patches on their computers :-)

http://wiki.sagemath.org/combinat/SageCombinatDaysCernay2012

Nathann

[jdemeyer]

Changed dependencies from 12235 to #12235

[jdemeyer]

comment:9

Please fix the commit message of the first patch.

[nathanncohen]

Attachment: trac_12306.patch.gz

[nathanncohen]

comment:10

Gloops... Done ! ^^;

Nathann

[jdemeyer]

Merged: sage-5.0.beta5