def test_perceptron(w, input_file, output_file):
with open(input_file) as f_in, open(output_file, 'w') as f_out:
for line in f_in:
sentence = line.rstrip()
phi = create_features(sentence)
prediction = predict_one(w, phi)
print(f'{prediction}\t{sentence}', file=f_out)
def test_perceptron(w, input_file):
with open(input_file, 'r', encoding='utf-8') as i_file:
for line in i_file:
snetence = line.rstrip()
phi = create_features(snetence)
prediction = predict_one(w, phi)
print(f'{prediction}\t{snetence}')
def predict_all(model_file, input_fle, n):
w = load_model(model_file)
result = []
for sentence in input_fle:
phi = defaultdict(lambda: 0)
phi = create_features(sentence.lower(), n, phi)
result.append(predict_one(w, phi))
return result
def predict_all(model_file, input_file):
with open(model_file, mode='rb') as model_f \
, open(input_file) as in_f:
weights = pickle.load(model_f)
#入力ファイルを一行ずつsplit
for line in in_f:
phi = create_features(line)
label_pred = predict_one(weights, phi)
print(label_pred)
def predict_all(model_file, input_file):
with open(model_file, 'r') as m_f, open(input_file, 'r') as i_f:
w = defaultdict(lambda: 0)
for line in m_f:
word, value = line.split('\t')
w[word] = int(value)
for x in i_f:
phi = defaultdict(lambda: 0)
for k, v in create_features(x).items():
phi[k] = int(v)
y_ = predict_one(w, phi)
yield ('{}\t{}'.format(y_, x))
def predict_all():
w = defaultdict(int)
with open('../../data/titles-en-test.word') as i_f, open(
'model_file.txt') as m_f:
for line in m_f:
words = line.strip().split()
w[words[0]] = int(words[1])
for x in i_f:
phi = create_features(x)
y_ = predict_one(w, phi)
printline = (str(y_) + '\t' + str(x)).strip('\n')
print(printline)
def pred_all(model_file, input_file):
phi = defaultdict(lambda : 0)
with open(model_file) as modelf:
w = defaultdict(lambda: 0)
for model_line in modelf:
name, weight = model_line.split('\t')
w[name] = int(weight)
with open(input_file) as inputf, open(ansfile, 'w') as ansout:
for in_line in inputf:
l = in_line
phi = create_features(l.lower())
y_pred = int(predict_one(w,phi))
print('{}\t{}'.format(y_pred, l), end='', file=ansout)
def test_create_features(self):
ansdict = {
'UNI:A': 1,
'UNI:site': 1,
'UNI:,': 2,
'UNI:Site': 1,
'UNI:in': 1,
'UNI:Maizuru': 1,
'UNI:Kyoto': 1
}
# Create features
outdict = create_features('A site , Site in Maizuru , Kyoto')
self.assertTrue(ansdict == outdict)
def test_perceptron(test_file, model_file, output_file):
'''パーセプトロンを用いた分類器テスト'''
# モデルの読み込み
w = defaultdict(float)
for line in open(model_file, encoding='utf8'):
key, raw_value = line.strip().split('\t')
w[key] = float(raw_value)
# テスト
with open(output_file, 'w', encoding='utf8') as f:
for line in open(test_file, encoding='utf8'):
raw_sentence = line.strip()
phi = create_features(raw_sentence)
prediction = predict_one(w, phi)
f.write(f'{prediction}\t{raw_sentence}\n')
def test_perceptron(model_file, test_file, output_file):
"""Predict all on the test file with a trained model.
(Described in Neubig's slides p.12.)
"""
w = load_model(model_file)
# Write the result into a buffer.
out = io.StringIO()
with open(test_file, 'r') as f:
for x in f:
phi = train_perceptron.create_features(x)
y_prime = train_perceptron.predict_one(w, phi)
out.write('{}\t{}\n'.format(y_prime, x.strip()))
# Print on the screen or save in the file.
if output_file == 'stdout':
print(out.getvalue().strip())
else:
with open(output_file, 'w') as f:
f.write(out.getvalue().strip())
# 重みを読み込み、予測を１行ずつ出力
from train_perceptron import create_features
from train_perceptron import predict_one
w = {} # word : weight
with open('model','r') as model:
for line in model:
columm= line.strip().split('\t')
w[columm[0]] = int(columm[1])
test_path = '../../data/titles-en-test.word' # テスト用
# test_path = 'test.txt'
with open(test_path,'r') as test_file:
for line in test_file:
line = line.rstrip()
phi = create_features(str(line))
predicted_label = predict_one(w, phi) # sign(w * phi)を計算
print(predicted_label)
# ../../script/grade-prediction.py data/titles-en-test.labeled result
from train_perceptron import create_features, predict_one
from collections import defaultdict
W = defaultdict(int)
for line in open('train_result.txt', 'r'):
key, value = line.split('\t')
W[key] = int(value)
for x in open('../../data/titles-en-test.word', 'r'):
phi = create_features(x)
y_prime = predict_one(W, phi)
print(y_prime)
import sys
from collections import defaultdict
from train_perceptron import create_features, predict_one
with open(sys.argv[1],
"r") as model_file, open(sys.argv[2], "r") as input_file, open(
"test_result.txt", "w") as fout:
w = defaultdict(int)
for line in model_file:
splited_line = line.strip().split("\t")
w[splited_line[0]] = int(splited_line[1])
for x in input_file:
phi = create_features(x.strip())
y_2 = predict_one(w, phi)
fout.write(f"{y_2}\n")
# Accuracy = 80.517180%