def benchmark_similarity(alpha,k,options,args):
fields = options.fields.split(" ")
dim = options.dim.split(" ")
types = options.types.split(" ")
#test()
#For csv
#"(\d\d\d\d-\d\d-\d\d \d\d:\d\d:\d\d),[^,]*,[^,]*,([^,]*),([^,]*),[^,]*,[^,]*,[^,]*,[^,]*,[^,]*,[^,]*,[^,]*,([^,]*)"
dict_dim = {}
if len(dim) != len(types):
raise Exception("The number of defined dimensions and types has to be equal")
else:
for i in range(len(dim)):
dict_dim[dim[i]] = types[i]
trees = build_aggregate_tree(options,args,fields,dim,types,dict_dim)
options.input = '../test/app_ipv4/random.%s.reverse.txt'%k
trees1 = build_aggregate_tree(options,args,fields,dim,types,dict_dim)
print "Computing Similarity"
s0 = maxSimilarity(trees[0][0],trees[-1][0])
print "still computing"
s1 = maxSimilarity(trees[0][0],trees1[0][0])
print "done"
return len(trees[0][0].preorder()),s0,s1
def test_edit_distance():
options,args = parse_options()
fields = options.fields.split(" ")
dim = options.dim.split(" ")
types = options.types.split(" ")
#test()
#For csv
#"(\d\d\d\d-\d\d-\d\d \d\d:\d\d:\d\d),[^,]*,[^,]*,([^,]*),([^,]*),[^,]*,[^,]*,[^,]*,[^,]*,[^,]*,[^,]*,[^,]*,([^,]*)"
dict_dim = {}
if len(dim) != len(types):
raise Exception("The number of defined dimensions and types has to be equal")
else:
for i in range(len(dim)):
dict_dim[dim[i]] = types[i]
fout = open("similarity.edit.%s.dat"%options.offset,"w")
upper_limit = options.batch + options.offset
for i in xrange(0,options.batch):
i = i + options.offset
lrow = "%s\t"%i
rrow = ""
for alpha in [0.02, 0.5,1,2,4]:
options.aggregate = alpha
options.input = '../test/app_ipv4/random.%s.txt'%i
trees = build_aggregate_tree(options,args,fields,dim,types,dict_dim)
t = T(trees[0][0].get_root())
options.input = '../test/app_ipv4/random.%s.reverse.txt'%i
trees1 = build_aggregate_tree(options,args,fields,dim,types,dict_dim)
t1 = T(trees1[0][0].get_root())
options.input = '../test/app_ipv4/random.%s.txt'%(upper_limit-i)
trees3 = build_aggregate_tree(options,args,fields,dim,types,dict_dim)
t3 = T(trees3[0][0].get_root())
print "Edit Distance Based Similarity"
s0 = compare.similarity(t,t1,g)
s1 = compare.similarity(t,t3,g)
lrow += "%s\t"%s0
rrow += "%s\t"%s1
fout.write(lrow+rrow+"\n")
fout.close()
def main():
sys.setrecursionlimit(1000000)
options,args = parse_options()
fields = options.fields.split(" ")
dim = options.dim.split(" ")
types = options.types.split(" ")
#test()
#For csv
#"(\d\d\d\d-\d\d-\d\d \d\d:\d\d:\d\d),[^,]*,[^,]*,([^,]*),([^,]*),[^,]*,[^,]*,[^,]*,[^,]*,[^,]*,[^,]*,[^,]*,([^,]*)"
dict_dim = {}
if len(dim) != len(types):
raise Exception("The number of defined dimensions and types has to be equal")
else:
for i in range(len(dim)):
dict_dim[dim[i]] = types[i]
fout = open("similarity.2010224.test.dat","w")
for alpa in [2]:
options.aggregate = alpa
options.input = '../test/app_ipv4/20100224/20100224.%s.txt'%i
trees = build_aggregate_tree(options,args,fields,dim,types,dict_dim)
t = T(trees[0][0].get_root())
options.input = '../test/app_ipv4/20100224/20100224.%s.reverse.txt'%i
trees2 = build_aggregate_tree(options,args,fields,dim,types,dict_dim)
t2 = T(trees2[0][0].get_root())
simil = compare.similarity(t,t2,g)