fix README

added TODO list

README.md

@@ -3,22 +3,22 @@
 This module implements the Markov Cluster algorithm created by Stijn van Dongen and 
 described in http://www.library.uu.nl/digiarchief/dip/diss/1895620/inhoud.htm.
 
-This implementation si not yet optimized for large networks.
+This implementation is not yet optimized for large networks.
 
 ![example] (example.png)
 
 
-## Installation:
+## Installation
 
 python setup.py install
 
-##Usage:
+##Usage
 
-###Command line:
+###Command line
 
     Usage: mcl_clustering.py [options] <input_matrix>
 
-    Options:
+    Options
       -h, --help            show this help message and exit
       -e EXPAND_FACTOR, --expand_factor=EXPAND_FACTOR
                             expand factor (default: 2)
@@ -41,7 +41,7 @@ An example is the file example.csv.
 try mcl_clustering.py -e 3 example.csv
 
 
-###Code:
+###Code
 
 ####Using numpy adjacency matrix
 
@@ -66,32 +66,30 @@ try mcl_clustering.py -e 3 example.csv
                        mult_factor = <mult_factor>)
 
 
-    Output:
+    Output
         M = otuput matrix
         clusters = dict with keys = [<cluster id>] values = [<vertex id>]
 
-##Requirements:
+##Requirements
 
     numpy
     networkx
 
-
-##Parameters:
+##Parameters
 
     -i --inflate-factor
     -e --expand-factor
     -m --multiply-factor
     -l --max-loops
     -s --show-graph     show graph with networkx
 
+## References
 
-## References:
-
-[1]   Stijn van Dongen, Graph Clustering by Flow Simulation.
+* [1]   Stijn van Dongen, Graph Clustering by Flow Simulation.
       PhD thesis, University of Utrecht, May 2000.
       ( http://www.library.uu.nl/digiarchief/dip/diss/1895620/inhoud.htm )
 
-[2]   Stijn van Dongen. A cluster algorithm for graphs.  Technical Report
+* [2]   Stijn van Dongen. A cluster algorithm for graphs.  Technical Report
       INS-R0010, National Research Institute for Mathematics and Computer
       Science in the Netherlands, Amsterdam, May 2000.
       ( http://www.cwi.nl/ftp/CWIreports/INS/INS-R0010.ps.Z )

TODO

@@ -0,0 +1,4 @@
+- improve graph visualization
+- improve unittest
+- use numpy sparse matrices instead of numpy arrays
+- speedup the algo

mcl/mcl_clustering.py

@@ -43,7 +43,7 @@ def draw(G, A, cluster_map):
 
     colors = []
     for i in range(len(G.nodes())):
-        colors.append( clust_map.get(i, 100 ))
+        colors.append(clust_map.get(i, 100))
 
     pos = nx.spring_layout(G)
     nx.draw_networkx_nodes(G, pos,node_size = 200, node_color =colors , cmap=plt.cm.Blues )
@@ -70,7 +70,6 @@ def mcl(M, expand_factor = 2, inflate_factor = 2, max_loop = 10 , mult_factor =
     M = add_diag(M, mult_factor)
     M = normalize(M)
 
-
     for i in range(max_loop):
         logging.info("loop", i)
         M = inflate(M, inflate_factor)
@@ -149,6 +148,16 @@ def get_graph(csv_filename):
     G = nx.from_numpy_matrix(np.matrix(M))
     return np.array(M), G
 
+def clusters_to_output(clusters, options):
+    if options.output and len(options.output)>0:
+        f = open(options.output, 'w')
+        for k, v in clusters.items():
+            f.write("%s|%s\n" % (k, ", ".join(map(str, v)) ))
+        f.close()
+    else:    
+        print "Clusters:"
+        for k, v in clusters.items():
+            print k, v
 
 if __name__ == '__main__':
 
@@ -165,15 +174,7 @@ def get_graph(csv_filename):
                                mult_factor = options.mult_factor)
     print time.time(), "done\n"
 
-    if options.output and len(options.output)>0:
-        f = open(options.output, 'w')
-        for k, v in clusters.items():
-            f.write("%s|%s\n" % (k, ", ".join(map(str, v)) ))
-        f.close()
-    else:    
-        print "Clusters:"
-        for k, v in clusters.items():
-            print k, v
+    clusters_to_output(clusters, options)
 
     if options.draw:
         print time.time(), "drawing..."

setup.py

@@ -3,9 +3,9 @@
 from distutils.core import setup
 
 setup(
-    name='MCL Markov Clustering',
+    name='MCL Markov Cluster',
     version='0.3',
-    description='Markov Clustering algoritm for Graphs',
+    description='Markov Cluster algorithm implementation',
     scripts = [
         'mcl/mcl_clustering.py'
     ],