counts[j][c] += 1
for i in xrange(len(strand)):
new += maxChar(counts[i])
return new
if __name__ == "__main__":
dnaFuncs = ks.kMeansInfo(distance, centroids)
k = 4
#(data, truths) = dna.makeStrands(k, 100, 5000, 10)
(data, truths) = dna.makeStrands(k, 100, 10000, 2)
#data = ['ACGT', 'CCTT', 'TCGT', 'CGTT', 'ACGG', 'ATGT']
"""
with open("benchmark_data.txt", "w") as fp:
fp.write(pickle.dumps(data))
"""
with open("benchmark_data.txt", "r") as fp:
data = pickle.loads(fp.read())
print hash(fp.read())
obj = ks.kMeansCluster(dnaFuncs, data, k, .01)
if (comm.Get_rank() == ROOT):
t1 = time.time()
res = obj.findClusters()
t2 = time.time()
#with open("foo.txt", "w") as fp:
# fp.write(str(res) + '\n\n' + str(t2 - t1))
print t2 - t1
else:
obj.findClusters()
for (x,y) in points:
x_sum += x
y_sum += y
final_x = x_sum / len(points)
final_y = y_sum / len(points)
return (final_x, final_y)
#points gives data in the form [(x1,y1), (x2,y2), ....]
if __name__ == "__main__":
pointFuncs = ks.kMeansInfo(distance, findCentroids)
data = [(-10,-10),(-11, -11),
(-10,10),(-11,11),
(10,-10),(11,-11),
(10,10),(11,11)]
obj = ks.kMeansCluster(pointFuncs, data, 4, .01)
if (comm.Get_rank() == ROOT):
print "Initial cluster"
print " ", data
res = obj.findClusters()
print "Final result:"
print " ", res
else:
obj.findClusters()