toxi-text-3M / preprocess_conversations.py
FredZhang7's picture
example of how to preprocess conversations
e0e5b16
raw
history blame
979 Bytes
import csv
# Open sexting_conversations.txt and read all lines
with open('sexting_conversations.txt', 'r', encoding='utf-8') as f:
lines = f.readlines()
# Initialize variables
data = []
text = ""
char_count = 0
# Loop through each line in the file
for line in lines:
# Increase character count
char_count += len(line)
# Add line to text
text += line
# 92 is the avg character length per line multiplied by 2
if char_count >= 92:
# Add text to data array with is_toxic value of 1
data.append([text.strip(), 1])
# Reset text and character count
text = ""
char_count = 0
# If there is any remaining text, add it to data array with is_toxic value of 1
if text:
data.append([text.strip(), 1])
# Save data array to CSV with column "text,is_toxic"
with open('output.csv', 'w', newline='', encoding='utf-8') as f:
writer = csv.writer(f)
writer.writerow(['text', 'is_toxic'])
writer.writerows(data)