def threshold_iterator(threshold):
"""
The method used to extract data from the XML-file and calculate distances.
Iterates over different thresholds to find the best threshold or just
calculates the entailment values for one threshold.
"""
global idf_dict
dom_doc = xml_util.get_dom_from_xml("data/RTE2_dev.preprocessed.xml")
pairs = xml_util.get_pairs(dom_doc)
pair_attributes = xml_util.get_attributes_from_preprocessed_pair_nodes(pairs)
idf_dict = calculate_idf_dictionary(pair_attributes)
print len(idf_dict.keys())
tree_value_pairs = []
# Extracting the actual lemma values from the pair nodes
for i in range(len(pair_attributes)):
t,h,id_num,e,ta = pair_attributes[i]
id_num = int(id_num)
t_values = xml_util.get_minipar_values_from_text_node(t)
h_values = xml_util.get_minipar_values_from_text_node(h)
tree_value_pairs.append((t_values,h_values))
# Calculating distances between text and hypothesis
distances = []
for i in range(len(tree_value_pairs)):
t_tree,h_tree = build_tree(tree_value_pairs[i])
dist = tree_edit_dist.distance(t_tree, h_tree, idf_cost)
normalizer = tree_edit_dist.distance(tree_edit_dist.Node("root"), h_tree, idf_cost)
normalized_dist = float(dist) / float(normalizer)
distances.append(normalized_dist)
#for d in distances:
# print d
if threshold == -1:
for i in range(200):
threshold = 1.0 - (0.005 * i)
syntax_matching(pair_attributes, distances, threshold)
else:
syntax_matching(pair_attributes, distances, threshold)
for node in soup.findAll("hypothesis"):
hypo.append(node.findAll("node"))
sentence = soup.findAll("node")
text = make_nodes(text[0])
hypo = make_nodes(hypo[0])
text = process_node(text)
hypo = process_node(hypo)
text_tree = make_tree(text)
hypo_tree = make_tree(hypo)
print hypo_tree
res = tree_edit_dist.distance(text_tree, hypo_tree)
print "distance =", res
##t1 = {"c":["b"],"d":["a","c"],"f":["d","e"], "b":[], "e":[], "a":[]}
##t2 = {"c":["d"],"d":["a","b"],"f":["c","e"], "b":[], "e":[], "a":[]}
##test_tree1 = recur_tree(t1["f"], t1, "f", [])
##test_tree2 = recur_tree(t2["f"], t2, "f", [])
#gg = recur_tree(d["Root"], d, "Root", [])
#print gg
#print test_tree1
#print test_tree2
#print gg
#for i in gg:
# print i[0],
# for j in i[1]:
