def main(src, dst, percent):
data = scan_data(src)
paths = []
for _, path in data:
paths.append(path)
tot = len(paths)
#shuffle
shuffle(paths)
count = 0
for p in paths[:int(percent * tot)]:
dst_p = p.replace(src, dst)
os.makedirs(dst_p[:dst_p.rindex(os.path.sep)], exist_ok=True)
copy(p, dst_p)
count += 1
print("Copied %d files of %d" % (count, tot))
# Topic : USC CSCI 544 Applied NLP Fall 16 - HW2 # - Perceptron Classifier for Spam-Ham classification # Author : Thamme Gowda Narayanaswamy # Student ID : 2074-6694-39 # Email : [email protected] # Date : Oct 12, 2016 import sys import os from perceptron import AvgPerceptron from perceptron import scan_data if __name__ == '__main__': train_dir = sys.argv[1] data = list(scan_data(train_dir)) print("Found %d total examples" % len(data)) model = AvgPerceptron() model.learn(data, 30) model.save('per_model.txt')
# Topic : USC CSCI 544 Applied NLP Fall 16 - HW2 # - Perceptron Classifier for Spam-Ham classification # Author : Thamme Gowda Narayanaswamy # Student ID : 2074-6694-39 # Email : [email protected] # Date : Oct 12, 2016 import sys import os from perceptron import StdPerceptron, scan_data if __name__ == '__main__': train_dir = sys.argv[1] data = list(scan_data(train_dir)) print("Found %d total examples" % len(data)) model = StdPerceptron() model.learn(data, 20) model.save('per_model.txt')
