def hierarchical_control():
print_line()
Printer.indent_right()
while True:
Printer.print("Select Vector Space Type:")
Printer.indent_right()
Printer.print("""1. Word2Vec Space
2. Tf-Idf Space
0. Back""")
Printer.indent_left()
inp1 = input()
if inp1 == '1':
print_line()
Printer.indent_right()
Printer.print("Word2Vec Space Selected")
output_list = []
while True:
print_line()
Printer.print("Enter Cluster Count: (Enter 0 for ending)")
try:
count = int(input())
except:
invalid()
continue
if count <= 0:
Printer.indent_left()
break
labels, model = cm.cluster_w2v_with_agglomerative_cluster(count)
Printer.print(f'Clustering Finished')
plt.title('Hierarchical Clustering Dendrogram')
plot_dendrogram(model, truncate_mode='lastp', p=count)
plt.show()
elif inp1 == '2':
print_line()
Printer.indent_right()
Printer.print("Tf-Idf Space Selected")
output_list = []
while True:
print_line()
Printer.print("Enter Cluster Count: (Enter 0 for ending)")
try:
count = int(input())
except:
invalid()
continue
if count <= 0:
Printer.indent_left()
break
labels, model = cm.cluster_tf_idf_with_agglomerative_cluster(count)
Printer.print(f'Clustering Finished')
plt.title('Hierarchical Clustering Dendrogram')
plot_dendrogram(model, truncate_mode='lastp', p=count)
plt.show()
elif inp1 == '0':
Printer.indent_left()
return
else:
invalid()
def invalid():
Printer.print("Invalid input!")
print_line()
def k_means_control():
print_line()
Printer.indent_right()
while True:
Printer.print("Select Vector Space Type:")
Printer.indent_right()
Printer.print("""1. Word2Vec Space
2. Tf-Idf Space
0. Back""")
Printer.indent_left()
inp1 = input()
if inp1 == '1':
print_line()
Printer.indent_right()
Printer.print("Word2Vec Space Selected")
output_list = []
while True:
print_line()
Printer.print("Enter Cluster Count: (Enter 0 for ending)")
try:
count = int(input())
except:
invalid()
continue
if count <= 0:
Printer.indent_left()
break
inertia, labels, distances = cm.cluster_w2v_with_k_means(count)
Printer.print(f'Clustering Finished, Inertia: {inertia}')
output_list.append((count, inertia))
output_list.sort(key=lambda x: x[0])
labels = []
values = []
for output in output_list:
labels.append(output[0])
values.append(output[1])
plt.title('K-Means Word2Vec Inertia Plot')
plt.plot(labels, values, linewidth=2, color='blue', marker='o')
plt.show()
elif inp1 == '2':
print_line()
Printer.indent_right()
Printer.print("Tf-Idf Space Selected")
output_list = []
while True:
print_line()
Printer.print("Enter Cluster Count: (Enter 0 for ending)")
try:
count = int(input())
except:
invalid()
continue
if count <= 0:
Printer.indent_left()
break
inertia, labels, distances = cm.cluster_tf_idf_with_k_means(count)
Printer.print(f'Clustering Finished, Inertia: {inertia}')
output_list.append((count, inertia))
output_list.sort(key=lambda x: x[0])
labels = []
values = []
for output in output_list:
labels.append(output[0])
values.append(output[1])
plt.title('K-Means Tf-Idf Inertia Plot')
plt.plot(labels, values, linewidth=2, color='blue', marker='o')
plt.show()
elif inp1 == '0':
Printer.indent_left()
return
else:
invalid()
def gaussian_control():
print_line()
Printer.indent_right()
while True:
Printer.print("Select Vector Space Type:")
Printer.indent_right()
Printer.print("""1. Word2Vec Space
2. Tf-Idf Space
0. Back""")
Printer.indent_left()
inp1 = input()
if inp1 == '1':
print_line()
Printer.indent_right()
Printer.print("Word2Vec Space Selected")
output_list = []
while True:
print_line()
Printer.print("Enter Cluster Count: (Enter 0 for ending)")
try:
count = int(input())
except:
invalid()
continue
if count <= 0:
Printer.indent_left()
break
score, labels = cm.cluster_w2v_with_gaussian_mixture(count)
total_score = np.sum(score)
Printer.print(f'Clustering Finished, Score: {total_score}')
output_list.append((count, total_score))
output_list.sort(key=lambda x: x[0])
labels = []
values = []
for output in output_list:
labels.append(output[0])
values.append(output[1])
plt.title('Gaussian Mixture Word2Vec Score Plot')
plt.plot(labels, values, linewidth=2, color='red', marker='o')
plt.show()
elif inp1 == '2':
print_line()
Printer.indent_right()
Printer.print("Tf-Idf Space Selected")
output_list = []
while True:
print_line()
Printer.print("Enter Cluster Count: (Enter 0 for ending)")
try:
count = int(input())
except:
invalid()
continue
if count <= 0:
Printer.indent_left()
break
score, labels = cm.cluster_tf_idf_with_gaussian_mixture(count)
total_score = np.sum(score)
Printer.print(f'Clustering Finished, Score: {total_score}')
output_list.append((count, total_score))
output_list.sort(key=lambda x: x[0])
labels = []
values = []
for output in output_list:
labels.append(output[0])
values.append(output[1])
plt.title('Gaussian Mixture Tf-Idf Score Plot')
plt.plot(labels, values, linewidth=2, color='red', marker='o')
plt.show()
elif inp1 == '0':
Printer.indent_left()
return
else:
invalid()
plot_dendrogram(model, truncate_mode='lastp', p=count)
plt.show()
elif inp1 == '0':
Printer.indent_left()
return
else:
invalid()
if __name__ == '__main__':
cm = ClusterManager()
cm.load_corpus('DataSet/corpus/Phase3_Data.csv', remove_mentions=True)
cm.load_w2v_model('DataSet/models/deps.words', binary=False)
cm.corpus_generate_tf_idf_model()
cm.corpus_generate_word2vec_model()
print_line()
while True:
Printer.print("Select Clustering Method:")
Printer.indent_right()
Printer.print("""1. K-Means
2. Gaussian Mixture
3. Hierarchical Clustering
0. Exit""")
Printer.indent_left()
selection = input()
if selection == '1':
k_means_control()
elif selection == '2':
gaussian_control()
elif selection == '3':