Spaces:
Runtime error
Runtime error
Update app.py
Browse files
app.py
CHANGED
|
@@ -2,9 +2,10 @@ import gradio as gr
|
|
| 2 |
import pandas as pd
|
| 3 |
from transformers import pipeline
|
| 4 |
import matplotlib.pyplot as plt
|
|
|
|
| 5 |
|
| 6 |
-
# Configurar el clasificador de sentimientos multiling眉e
|
| 7 |
-
classifier = pipeline(task="zero-shot-classification", model="
|
| 8 |
|
| 9 |
# Funci贸n para analizar los sentimientos de una lista de textos
|
| 10 |
def analyze_sentiments(texts):
|
|
@@ -12,7 +13,10 @@ def analyze_sentiments(texts):
|
|
| 12 |
return "0.0%", "0.0%", "0.0%", None # Manejar el caso donde no hay textos para analizar
|
| 13 |
|
| 14 |
positive, negative, neutral = 0, 0, 0
|
| 15 |
-
|
|
|
|
|
|
|
|
|
|
| 16 |
results = classifier(text, ["positive", "negative", "neutral"], multi_label=True)
|
| 17 |
mx = max(results['scores'])
|
| 18 |
ind = results['scores'].index(mx)
|
|
@@ -23,11 +27,16 @@ def analyze_sentiments(texts):
|
|
| 23 |
negative += 1
|
| 24 |
else:
|
| 25 |
neutral += 1
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 26 |
total = len(texts)
|
| 27 |
positive_percent = round((positive / total) * 100, 1)
|
| 28 |
negative_percent = round((negative / total) * 100, 1)
|
| 29 |
neutral_percent = round((neutral / total) * 100, 1)
|
| 30 |
-
|
| 31 |
# Crear el gr谩fico circular
|
| 32 |
fig, ax = plt.subplots()
|
| 33 |
ax.pie([positive_percent, negative_percent, neutral_percent], labels=["Positivo", "Negativo", "Neutro"], autopct='%1.1f%%', colors=['green', 'red', 'blue'])
|
|
@@ -37,13 +46,13 @@ def analyze_sentiments(texts):
|
|
| 37 |
|
| 38 |
return f"{positive_percent}%", f"{negative_percent}%", f"{neutral_percent}%", "sentiment_pie_chart.png"
|
| 39 |
|
| 40 |
-
# Funci贸n para cargar el archivo CSV y analizar los primeros
|
| 41 |
def analyze_sentiment_from_csv(file):
|
| 42 |
try:
|
| 43 |
df = pd.read_csv(file.name)
|
| 44 |
if 'content' not in df.columns:
|
| 45 |
raise ValueError("El archivo CSV no contiene una columna 'content'")
|
| 46 |
-
texts = df['content'].head(
|
| 47 |
return analyze_sentiments(texts)
|
| 48 |
except pd.errors.ParserError as e:
|
| 49 |
return f"Error al analizar el archivo CSV: {e}", "", "", None
|
|
@@ -60,9 +69,10 @@ demo = gr.Interface(
|
|
| 60 |
gr.Textbox(label="Porcentaje Neutro"),
|
| 61 |
gr.Image(type="filepath", label="Gr谩fico de Sentimientos")
|
| 62 |
],
|
| 63 |
-
title="Analizador de Sentimientos V.
|
| 64 |
-
description="Porcentaje de comentarios positivos, negativos y neutrales
|
| 65 |
)
|
| 66 |
|
| 67 |
demo.launch(share=True)
|
| 68 |
|
|
|
|
|
|
| 2 |
import pandas as pd
|
| 3 |
from transformers import pipeline
|
| 4 |
import matplotlib.pyplot as plt
|
| 5 |
+
from concurrent.futures import ThreadPoolExecutor
|
| 6 |
|
| 7 |
+
# Configurar el clasificador de sentimientos multiling眉e con un modelo m谩s peque帽o
|
| 8 |
+
classifier = pipeline(task="zero-shot-classification", model="typeform/distilbert-base-uncased-mnli")
|
| 9 |
|
| 10 |
# Funci贸n para analizar los sentimientos de una lista de textos
|
| 11 |
def analyze_sentiments(texts):
|
|
|
|
| 13 |
return "0.0%", "0.0%", "0.0%", None # Manejar el caso donde no hay textos para analizar
|
| 14 |
|
| 15 |
positive, negative, neutral = 0, 0, 0
|
| 16 |
+
|
| 17 |
+
# Funci贸n para procesar un texto individualmente
|
| 18 |
+
def process_text(text):
|
| 19 |
+
nonlocal positive, negative, neutral
|
| 20 |
results = classifier(text, ["positive", "negative", "neutral"], multi_label=True)
|
| 21 |
mx = max(results['scores'])
|
| 22 |
ind = results['scores'].index(mx)
|
|
|
|
| 27 |
negative += 1
|
| 28 |
else:
|
| 29 |
neutral += 1
|
| 30 |
+
|
| 31 |
+
# Usar ThreadPoolExecutor para procesar textos en paralelo
|
| 32 |
+
with ThreadPoolExecutor() as executor:
|
| 33 |
+
executor.map(process_text, texts)
|
| 34 |
+
|
| 35 |
total = len(texts)
|
| 36 |
positive_percent = round((positive / total) * 100, 1)
|
| 37 |
negative_percent = round((negative / total) * 100, 1)
|
| 38 |
neutral_percent = round((neutral / total) * 100, 1)
|
| 39 |
+
|
| 40 |
# Crear el gr谩fico circular
|
| 41 |
fig, ax = plt.subplots()
|
| 42 |
ax.pie([positive_percent, negative_percent, neutral_percent], labels=["Positivo", "Negativo", "Neutro"], autopct='%1.1f%%', colors=['green', 'red', 'blue'])
|
|
|
|
| 46 |
|
| 47 |
return f"{positive_percent}%", f"{negative_percent}%", f"{neutral_percent}%", "sentiment_pie_chart.png"
|
| 48 |
|
| 49 |
+
# Funci贸n para cargar el archivo CSV y analizar los primeros 50 comentarios
|
| 50 |
def analyze_sentiment_from_csv(file):
|
| 51 |
try:
|
| 52 |
df = pd.read_csv(file.name)
|
| 53 |
if 'content' not in df.columns:
|
| 54 |
raise ValueError("El archivo CSV no contiene una columna 'content'")
|
| 55 |
+
texts = df['content'].head(50).tolist() # Tomar solo los primeros 50 comentarios
|
| 56 |
return analyze_sentiments(texts)
|
| 57 |
except pd.errors.ParserError as e:
|
| 58 |
return f"Error al analizar el archivo CSV: {e}", "", "", None
|
|
|
|
| 69 |
gr.Textbox(label="Porcentaje Neutro"),
|
| 70 |
gr.Image(type="filepath", label="Gr谩fico de Sentimientos")
|
| 71 |
],
|
| 72 |
+
title="Analizador de Sentimientos V.2",
|
| 73 |
+
description="Porcentaje de comentarios positivos, negativos y neutrales"
|
| 74 |
)
|
| 75 |
|
| 76 |
demo.launch(share=True)
|
| 77 |
|
| 78 |
+
|