from parse_debates import DebateParser
import ngrams
import os
import mallet
import build_graph
import networkx
'''
main entry point for generating mallet input files project
'''
if __name__ == "__main__":
# first parse and load the debates and create the mallet raw input
parser = DebateParser("./data/debates")
parser.parse()
parser.build_text_for_mallet('./data/mallet_raw_statements.txt')
# now generate and save bigrams for mallet replacement files
ngrams.save_bigrams_for_replacement_file_txt([item[0] for sublist in parser.statements.values() for item in sublist],
os.path.join("./data/mallet_files", "replacements.txt"))
with open(path, "wb+") as output_file:
writer = csv.writer(output_file, delimiter="\t")
if len(statements) > 0:
statements = [removeNonAscii(statement) for statement in statements]
ngrams = get_bigram_likelihood(statements)
if ngrams != '':
for ngram in ngrams:
writer.writerow([ngram[0][0] + '_' + ngram[0][1], ngram[1]])
def save_bigrams_for_replacement_file_txt(statements, path):
"""
saves likely bigrams in a txt
:param statements: list of strings
:param path: output path for saved txt
"""
with open(path, "wb+") as output_file:
if len(statements) > 0:
statements = [removeNonAscii(statement) for statement in statements]
ngrams = get_bigram_likelihood(statements)
if ngrams != '':
for ngram in ngrams:
output_file.write(ngram[0][0] + ' ' + ngram[0][1] + '\n')
if __name__ == "__main__":
parser = DebateParser("./data/debates")
parser.parse()
save_bigram_likelihood_tsv([item[0] for sublist in parser.statements.values() for item in sublist],
os.path.join("data", "ngrams.tsv"))
save_bigrams_for_replacement_file_txt([item[0] for sublist in parser.statements.values() for item in sublist],
os.path.join("./data/mallet_files", "replacements.txt"))
for ngram in ngrams:
writer.writerow(
[ngram[0][0] + '_' + ngram[0][1], ngram[1]])
def save_bigrams_for_replacement_file_txt(statements, path):
"""
saves likely bigrams in a txt
:param statements: list of strings
:param path: output path for saved txt
"""
with open(path, "wb+") as output_file:
if len(statements) > 0:
statements = [
removeNonAscii(statement) for statement in statements
]
ngrams = get_bigram_likelihood(statements)
if ngrams != '':
for ngram in ngrams:
output_file.write(ngram[0][0] + ' ' + ngram[0][1] + '\n')
if __name__ == "__main__":
parser = DebateParser("./data/debates")
parser.parse()
save_bigram_likelihood_tsv([
item[0] for sublist in parser.statements.values() for item in sublist
], os.path.join("data", "ngrams.tsv"))
save_bigrams_for_replacement_file_txt([
item[0] for sublist in parser.statements.values() for item in sublist
], os.path.join("./data/mallet_files", "replacements.txt"))
from parse_debates import DebateParser
import ngrams
import os
import mallet
import build_graph
import networkx
'''
main entry point for generating mallet input files project
'''
if __name__ == "__main__":
# first parse and load the debates and create the mallet raw input
parser = DebateParser("./data/debates")
parser.parse()
parser.build_text_for_mallet('./data/mallet_raw_statements.txt')
# now generate and save bigrams for mallet replacement files
ngrams.save_bigrams_for_replacement_file_txt([
item[0] for sublist in parser.statements.values() for item in sublist
], os.path.join("./data/mallet_files", "replacements.txt"))
