import time
google = NGD()
msn = NMD()
yahoo = NYD()
def compare( a, b , google, msn, yahoo):
g = google.distance(a,b)
m = msn.distance(a,b)
y = yahoo.distance(a,b)
print 'for "%s" "%s"' % (a,b)
print 'google: %f msn: %f yahoo: %f' % (g,m,y)
print ''
pairs = [('by','with'), ('quantum','physics'), ('quantum', 'football')]
print time.ctime()
b = time.time()
#print google.distances((pairs*30)[:21])
a = time.time()
print time.ctime()
print 'took %d seconds' % (a-b)
b = time.time()
print google.distances((pairs*30)[:65], True)
a = time.time()
print time.ctime()
print 'took %d seconds' % (a-b)
"article",
"paper",
]
# now we calculate the distance
name_best_match = {}
for n1 in mails_per_name.keys():
best_match_dist = 3.0
best_match = ("", "")
for context in contexts:
tuples = []
n2s = []
for n2 in mails_per_name.keys():
if n1 != n2:
tuples.append((n1, n2))
n2s.append(n2)
dists = ngd.distances(tuples, context)
for (n1, n2), dist in dists.iteritems():
if dist < best_match_dist:
best_match_dist = dist
best_match = (n2, context)
name_best_match[n1] = best_match
print "------------------------"
print "best_matches"
for n, (n2, context) in name_best_match.iteritems():
if n2 != "":
# use the non-domain email with the best hits
best = ""
best_hits = -1
for m in mails_per_name[n2]:
