def learn_triplets_cooccur_mat(file_in, co_mat_file):
learned_co_mat = CooccurMatrix()
learned_co_mat.load(co_mat_file)
np_voc = learned_co_mat.vocabulary
np1_matrix = learned_co_mat.matrix
np1_all = vocabulary.Vocabulary()
with open(file_in, 'r') as f:
for line in f:
if (line[0] != '<'):
line = (line[:-2]).lower()
triplets = line.split('|')
np1 = cleansing.clean(triplets[0].split())
# Delete words not in the similarity vocabulary.
np1_new = [w for w in np1 if np_voc.contain(w)]
for w in np1_new:
np1_all.add(w)
num_np1 = np1_all.size()
similarity_mat_np1 = zeros([num_np1, num_np1])
for i in range(num_np1):
for j in range(num_np1):
similarity_mat_np1[i, j] = np1_matrix[
np_voc.get_word_index(np1_all.get_word(i)),
np_voc.get_word_index(np1_all.get_word(j))]
return CooccurMatrix(similarity_mat_np1, np1_all)
def learn_triplets_cooccur_mat(file_in, co_mat_file):
learned_co_mat = CooccurMatrix()
learned_co_mat.load(co_mat_file)
np_voc = learned_co_mat.vocabulary
np1_matrix = learned_co_mat.matrix
np1_all = vocabulary.Vocabulary()
with open(file_in, 'r') as f:
for line in f:
if(line[0] != '<'):
line = (line[:-2]).lower()
triplets = line.split('|')
np1 = cleansing.clean(triplets[0].split())
# Delete words not in the similarity vocabulary.
np1_new = [w for w in np1 if np_voc.contain(w)]
for w in np1_new:
np1_all.add(w)
num_np1 = np1_all.size()
similarity_mat_np1 = zeros([num_np1, num_np1])
for i in range(num_np1):
for j in range(num_np1):
similarity_mat_np1[i, j] = np1_matrix[np_voc.get_word_index(np1_all.get_word(i)), np_voc.get_word_index(np1_all.get_word(j))]
return CooccurMatrix(similarity_mat_np1, np1_all)
def learn_story_distances(triplets_file_path, co_mat_file, use_similarity=True, min_similarity=None, output_file=False):
# Load learned similarity matrix.
learned_co_mat = CooccurMatrix()
learned_co_mat.load(co_mat_file)
np_voc = learned_co_mat.vocabulary
np1_matrix = learned_co_mat.matrix
if min_similarity is not None:
np1_matrix = (np1_matrix >= 0.8) * np1_matrix
files = glob.glob(triplets_file_path)
files.sort()
print len(files)
file_num = len(files)
# Calculate word histogram for each story.
np_num = np_voc.size()
hist = np.zeros([file_num, np_num])
count = 0
for file_in in files:
(h, wordlist) = learn_story_histogram(file_in, np_voc)
hist[count, :] = h
count += 1
# Calculate pair-wise distance between stories.
dist = np.zeros([file_num, file_num])
for i in range(file_num):
for j in range(file_num):
dif = hist[i, :] - hist[j, :]
if use_similarity:
sq = np.dot(np.dot(dif, np1_matrix), dif.T)
if (sq < 0):
sq = 0
dist[i, j] = sqrt(sq)
else:
dist[i, j] = sqrt(np.dot(dif, dif.T))
labels = []
for filename in files:
labels.append(filename.split('/')[-1][:-4])
if output_file:
np.savetxt('../mat/histogram.txt', hist)
np.savetxt('../mat/distance.txt', dist)
with codecs.open('../mat/filename.txt', "w", encoding='ISO-8859-1') as f:
for l in labels:
f.writelines(l + '\n')
return (dist, labels)
def learn_story_distances(triplets_file_path, co_mat_file, use_similarity=True, min_similarity=None, output_file=False):
# Load learned similarity matrix.
learned_co_mat = CooccurMatrix()
learned_co_mat.load(co_mat_file)
np_voc = learned_co_mat.vocabulary
np1_matrix = learned_co_mat.matrix
if min_similarity is not None:
np1_matrix = (np1_matrix >= 0.8) * np1_matrix
files = glob.glob(triplets_file_path)
files.sort()
print len(files)
file_num = len(files)
# Calculate word histogram for each story.
np_num = np_voc.size()
hist = np.zeros([file_num, np_num])
count = 0
for file_in in files:
(h, wordlist) = learn_story_histogram(file_in, np_voc)
hist[count, :] = h
count += 1
# Calculate pair-wise distance between stories.
dist = np.zeros([file_num, file_num])
for i in range(file_num):
for j in range(file_num):
dif = hist[i, :] - hist[j, :]
if use_similarity:
sq = np.dot(np.dot(dif, np1_matrix), dif.T)
if sq < 0:
sq = 0
dist[i, j] = sqrt(sq)
else:
dist[i, j] = sqrt(np.dot(dif, dif.T))
labels = []
for filename in files:
labels.append(filename.split("/")[-1][:-4])
if output_file:
np.savetxt("../mat/histogram.txt", hist)
np.savetxt("../mat/distance.txt", dist)
with codecs.open("../mat/filename.txt", "w", encoding="ISO-8859-1") as f:
for l in labels:
f.writelines(l + "\n")
return (dist, labels)
