35 lines
955 B
Python
35 lines
955 B
Python
|
from dataclasses import dataclass
|
||
|
|
from random import sample
|
||
|
|
from typing import List
|
||
|
|
|
||
|
|
import pandas as pd
|
||
|
|
|
||
|
|
from sawaw import SAWAWEntry, analyze_gpt3_5
|
||
|
|
|
||
|
|
# Create a DataFrame from ./combined_data_with_aspects.csv
|
||
|
|
df = pd.read_csv("data/combined_data_with_aspects.csv")
|
||
|
|
# Keep only text and aspect columns
|
||
|
|
df = df[["text", "aspect"]]
|
||
|
|
# Remove rows with NaN values
|
||
|
|
df = df.dropna()
|
||
|
|
# Remove rows with empty strings
|
||
|
|
df = df[df["text"] != ""]
|
||
|
|
# Create a list of SAWAWEntry objects
|
||
|
|
entries = []
|
||
|
|
for index, row in df.iterrows():
|
||
|
|
comment = row["text"]
|
||
|
|
aspect_words = row["aspect"].split(",")
|
||
|
|
aspect_words = [aspect_word.strip() for aspect_word in aspect_words]
|
||
|
|
entries.append(SAWAWEntry(comment, aspect_words))
|
||
|
|
|
||
|
|
# Pick 10 random entries
|
||
|
|
entries = sample(entries, 3)
|
||
|
|
# Query the entries
|
||
|
|
for entry in entries:
|
||
|
|
print(entry)
|
||
|
|
analyze_gpt3_5(entry)
|
||
|
|
SAWAWEntry.print_legends()
|
||
|
|
print(entry)
|
||
|
|
input("Press Enter to continue...")
|
||
|
|
|