def kruskal_MST(gr):
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 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 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(self):
self.uf = UnionFind()
self.uf.insert("a", "b")
self.uf.insert("b", "c")
self.uf.insert("i", "j")
class test_unionfind(unittest.TestCase):
def setUp(self):
self.uf = UnionFind()
self.uf.insert("a", "b")
self.uf.insert("b", "c")
self.uf.insert("i", "j")
def test_get_parent_method(self):
self.assertEqual("a", self.uf.get_leader("a"))
self.assertEqual("a", self.uf.get_leader("b"))
self.assertEqual("a", self.uf.get_leader("c"))
self.assertEqual("i", self.uf.get_leader("j"))
self.assertEqual("i", self.uf.get_leader("i"))
self.assertNotEqual(self.uf.get_leader("a"), self.uf.get_leader("i"))
def test_insert_method(self):
self.uf.insert("c", "d")
self.assertEqual(self.uf.get_leader("c"), self.uf.get_leader("d"))
self.assertEqual(self.uf.get_leader("a"), self.uf.get_leader("d"))
def test_make_union_method(self):
self.uf.make_union(self.uf.get_leader("a"), self.uf.get_leader("i"))
self.assertEqual(self.uf.get_leader("a"), self.uf.get_leader("i"))
def test_make_union_with_invalid_leader_raises_exception(self):
self.assertRaises(Exception, self.uf.make_union, "a", "z")
def setup_uf():
uf = UnionFind()
uf.insert("a", "b")
uf.insert("b", "c")
uf.insert("i", "j")
return uf
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 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 setUp(self):
self.uf = UnionFind()
self.uf.insert("a", "b")
self.uf.insert("b", "c")
self.uf.insert("i", "j")
class test_unionfind(unittest.TestCase):
def setUp(self):
self.uf = UnionFind()
self.uf.insert("a", "b")
self.uf.insert("b", "c")
self.uf.insert("i", "j")
def test_get_parent_method(self):
self.assertEqual("a", self.uf.get_leader("a"))
self.assertEqual("a", self.uf.get_leader("b"))
self.assertEqual("a", self.uf.get_leader("c"))
self.assertEqual("i", self.uf.get_leader("j"))
self.assertEqual("i", self.uf.get_leader("i"))
self.assertNotEqual(self.uf.get_leader("a"), self.uf.get_leader("i"))
def test_insert_method(self):
self.uf.insert("c", "d")
self.assertEqual(self.uf.get_leader("c"), self.uf.get_leader("d"))
self.assertEqual(self.uf.get_leader("a"), self.uf.get_leader("d"))
def test_insert_one_node(self):
self.uf.insert('z')
self.assertEqual(self.uf.get_leader('z'), 'z')
self.assertEqual(self.uf.count_groups(), 3)
def test_make_union_method(self):
self.uf.make_union(self.uf.get_leader("a"), self.uf.get_leader("i"))
self.assertEqual(self.uf.get_leader("a"), self.uf.get_leader("i"))
def test_make_union_with_invalid_leader_raises_exception(self):
self.assertRaises(Exception, self.uf.make_union, "a", "z")
def test_get_count(self):
self.uf.insert("z", "y")
self.assertEqual(self.uf.count_groups(), 3)
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]))
