Spaces:
Sleeping
Sleeping
Commit ·
55c5135
1
Parent(s): 6177fd7
progress more 42+
Browse files- app.py +10 -2
- sentiment_decorators.py +21 -0
app.py
CHANGED
|
@@ -13,6 +13,7 @@ import torch
|
|
| 13 |
from openpyxl import load_workbook
|
| 14 |
from openpyxl import Workbook
|
| 15 |
from openpyxl.utils.dataframe import dataframe_to_rows
|
|
|
|
| 16 |
|
| 17 |
# Initialize pymystem3 for lemmatization
|
| 18 |
mystem = Mystem()
|
|
@@ -85,22 +86,27 @@ def get_mapped_sentiment(result):
|
|
| 85 |
return "Negative"
|
| 86 |
return "Neutral"
|
| 87 |
|
|
|
|
| 88 |
def get_rubert1_sentiment(text):
|
| 89 |
result = rubert1(text, truncation=True, max_length=512)[0]
|
| 90 |
return get_mapped_sentiment(result)
|
| 91 |
|
|
|
|
| 92 |
def get_rubert2_sentiment(text):
|
| 93 |
result = rubert2(text, truncation=True, max_length=512)[0]
|
| 94 |
return get_mapped_sentiment(result)
|
| 95 |
|
|
|
|
| 96 |
def get_finbert_sentiment(text):
|
| 97 |
result = finbert(text, truncation=True, max_length=512)[0]
|
| 98 |
return get_mapped_sentiment(result)
|
| 99 |
|
|
|
|
| 100 |
def get_roberta_sentiment(text):
|
| 101 |
result = roberta(text, truncation=True, max_length=512)[0]
|
| 102 |
return get_mapped_sentiment(result)
|
| 103 |
|
|
|
|
| 104 |
def get_finbert_tone_sentiment(text):
|
| 105 |
result = finbert_tone(text, truncation=True, max_length=512)[0]
|
| 106 |
return get_mapped_sentiment(result)
|
|
@@ -149,7 +155,9 @@ def process_file(uploaded_file):
|
|
| 149 |
total_news = len(df)
|
| 150 |
|
| 151 |
texts = df['Выдержки из текста'].tolist()
|
| 152 |
-
|
|
|
|
|
|
|
| 153 |
for text in df['Выдержки из текста']:
|
| 154 |
lemmatized_texts.append(lemmatize_text(text))
|
| 155 |
|
|
@@ -249,7 +257,7 @@ def create_output_file(df, uploaded_file, analysis_df):
|
|
| 249 |
return output
|
| 250 |
|
| 251 |
def main():
|
| 252 |
-
st.title("... приступим к анализу... версия
|
| 253 |
|
| 254 |
uploaded_file = st.file_uploader("Выбирайте Excel-файл", type="xlsx")
|
| 255 |
|
|
|
|
| 13 |
from openpyxl import load_workbook
|
| 14 |
from openpyxl import Workbook
|
| 15 |
from openpyxl.utils.dataframe import dataframe_to_rows
|
| 16 |
+
from sentiment_decorators import sentiment_analysis_decorator
|
| 17 |
|
| 18 |
# Initialize pymystem3 for lemmatization
|
| 19 |
mystem = Mystem()
|
|
|
|
| 86 |
return "Negative"
|
| 87 |
return "Neutral"
|
| 88 |
|
| 89 |
+
@sentiment_analysis_decorator
|
| 90 |
def get_rubert1_sentiment(text):
|
| 91 |
result = rubert1(text, truncation=True, max_length=512)[0]
|
| 92 |
return get_mapped_sentiment(result)
|
| 93 |
|
| 94 |
+
@sentiment_analysis_decorator
|
| 95 |
def get_rubert2_sentiment(text):
|
| 96 |
result = rubert2(text, truncation=True, max_length=512)[0]
|
| 97 |
return get_mapped_sentiment(result)
|
| 98 |
|
| 99 |
+
@sentiment_analysis_decorator
|
| 100 |
def get_finbert_sentiment(text):
|
| 101 |
result = finbert(text, truncation=True, max_length=512)[0]
|
| 102 |
return get_mapped_sentiment(result)
|
| 103 |
|
| 104 |
+
@sentiment_analysis_decorator
|
| 105 |
def get_roberta_sentiment(text):
|
| 106 |
result = roberta(text, truncation=True, max_length=512)[0]
|
| 107 |
return get_mapped_sentiment(result)
|
| 108 |
|
| 109 |
+
@sentiment_analysis_decorator
|
| 110 |
def get_finbert_tone_sentiment(text):
|
| 111 |
result = finbert_tone(text, truncation=True, max_length=512)[0]
|
| 112 |
return get_mapped_sentiment(result)
|
|
|
|
| 155 |
total_news = len(df)
|
| 156 |
|
| 157 |
texts = df['Выдержки из текста'].tolist()
|
| 158 |
+
# Data validation
|
| 159 |
+
texts = [str(text) if not pd.isna(text) else "" for text in texts]
|
| 160 |
+
|
| 161 |
for text in df['Выдержки из текста']:
|
| 162 |
lemmatized_texts.append(lemmatize_text(text))
|
| 163 |
|
|
|
|
| 257 |
return output
|
| 258 |
|
| 259 |
def main():
|
| 260 |
+
st.title("... приступим к анализу... версия 42+")
|
| 261 |
|
| 262 |
uploaded_file = st.file_uploader("Выбирайте Excel-файл", type="xlsx")
|
| 263 |
|
sentiment_decorators.py
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import functools
|
| 2 |
+
from typing import Callable, Any
|
| 3 |
+
|
| 4 |
+
|
| 5 |
+
def sentiment_analysis_decorator(func: Callable[..., Any]) -> Callable[..., Any]:
|
| 6 |
+
@functools.wraps(func)
|
| 7 |
+
def wrapper(text: Any, *args: Any, **kwargs: Any) -> str:
|
| 8 |
+
if not isinstance(text, str):
|
| 9 |
+
if pd.isna(text):
|
| 10 |
+
return "Neutral" # nothing meanz neutral
|
| 11 |
+
text = str(text) # Convert to string
|
| 12 |
+
|
| 13 |
+
try:
|
| 14 |
+
result = func(text, *args, **kwargs)
|
| 15 |
+
return result
|
| 16 |
+
except Exception as e:
|
| 17 |
+
print(f"Error in {func.__name__} processing text: {text[:100]}...") # expose 100 chars of problematic text
|
| 18 |
+
print(f"Error: {str(e)}")
|
| 19 |
+
return "Neutral" # nothing meanz neutral
|
| 20 |
+
|
| 21 |
+
return wrapper
|