def max_k_clustering(gr, k):
sorted_edges = sorted(gr.get_edge_weights())
uf = UnionFind()
#initialize each node as its cluster
for n in gr.nodes():
uf.insert(n)
for (w, (u, v)) in sorted_edges:
if uf.count_groups() <= k:
return uf.get_sets()
if uf.get_leader(u) != uf.get_leader(v):
uf.make_union(uf.get_leader(u), uf.get_leader(v))
def kruskal_MST(gr):
""" computes minimum cost spanning tree in a undirected,
connected graph using Kruskal's MST. Uses union-find data structure
for running times of O(mlogn) """
sorted_edges = sorted(gr.get_edge_weights())
uf = UnionFind()
min_cost = 0
for (w, (u, v)) in sorted_edges:
if (not uf.get_leader(u) and not uf.get_leader(v)) \
or (uf.get_leader(u) != uf.get_leader(v)):
uf.insert(u, v)
min_cost += w
return min_cost
def is_graph_cyclic(gr):
uf = UnionFind()
edges_explored = set()
for node in gr.nodes():
uf.insert(node, node)
for edge in gr.edges():
if edge in edges_explored:
continue
node, neighbor = edge
edges_explored.add(edge)
edges_explored.add((neighbor, node))
x = uf.get_leader(node)
y = uf.get_leader(neighbor)
if x == y:
return True
uf.make_union(x, y)
return False
def setup_uf():
uf = UnionFind()
uf.insert("a", "b")
uf.insert("b", "c")
uf.insert("i", "j")
return uf
def setUp(self):
self.uf = UnionFind()
self.uf.insert("a", "b")
self.uf.insert("b", "c")
self.uf.insert("i", "j")
import os, sys
import operator
sys.path.append(os.path.join(os.getcwd(), os.path.pardir))
from graphs.graph import graph
from itertools import *
from union_find.unionfind import UnionFind
def ham_dist(e1, e2):
""" computes hamming distance between two strings e1 and e2 """
ne = operator.ne
return sum(imap(ne, e1, e2))
path = "clustering3.txt"
nodes = open(path).readlines()
uf = UnionFind()
# bitcount = {i: nodes[i].count('1') for i in range(len(nodes))}
# similar = [(bitcount[i]-9, ham_dist(nodes[i], nodes[0])) for i in range(1, len(nodes))]
# print nodes[1].count('1') - nodes[2].count('1')
# print hamdist(nodes[1], nodes[2])
for i in range(len(nodes)):
for j in range(i + 1, len(nodes)):
print(i, j, ham_dist(nodes[i], nodes[j]))
