Spaces:
Sleeping
Sleeping
Create app.py
Browse files
app.py
ADDED
|
@@ -0,0 +1,143 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import torch
|
| 2 |
+
import gradio as gr
|
| 3 |
+
import pandas as pd
|
| 4 |
+
import matplotlib.pyplot as plt
|
| 5 |
+
import seaborn as sns
|
| 6 |
+
|
| 7 |
+
|
| 8 |
+
# Use a pipeline as a high-level helper
|
| 9 |
+
from transformers import pipeline
|
| 10 |
+
|
| 11 |
+
# model_path = ("../Models/models--distilbert--distilbert-base-uncased-finetuned-sst-2-english/snapshots/714eb0fa89d2f80546fda750413ed43d93601a13")
|
| 12 |
+
# analyzer = pipeline("text-classification", model=model_path)
|
| 13 |
+
|
| 14 |
+
analyzer = pipeline("text-classification", model="distilbert/distilbert-base-uncased-finetuned-sst-2-english")
|
| 15 |
+
|
| 16 |
+
# print(analyzer(["This is good", "This product is quite expensive"]))
|
| 17 |
+
|
| 18 |
+
def sentiment_analyzer(review):
|
| 19 |
+
sentiment = analyzer(review)
|
| 20 |
+
return sentiment[0]['label'], sentiment[0]['score'] # Return sentiment label and confidence score
|
| 21 |
+
|
| 22 |
+
def create_sentiment_pie_chart(df):
|
| 23 |
+
sentiment_counts = df['Sentiment'].value_counts()
|
| 24 |
+
|
| 25 |
+
# Style the pie chart
|
| 26 |
+
colors = ['#e67e22', '#3498db'] # Green for Positive, Red for Negative
|
| 27 |
+
explode = [0.05, 0] # Slightly separate the positive slice for emphasis
|
| 28 |
+
|
| 29 |
+
fig, ax = plt.subplots(figsize=(6, 6))
|
| 30 |
+
sentiment_counts.plot(
|
| 31 |
+
kind='pie',
|
| 32 |
+
colors=colors,
|
| 33 |
+
autopct='%1.1f%%',
|
| 34 |
+
startangle=140,
|
| 35 |
+
explode=explode,
|
| 36 |
+
ax=ax,
|
| 37 |
+
textprops={'fontsize': 10, 'color': 'black'},
|
| 38 |
+
radius=0.7 # Decrease radius to make the pie smaller
|
| 39 |
+
)
|
| 40 |
+
ax.set_title("Sentiment Distribution (Pie Chart)", fontsize=10, fontweight='bold', color='#34495e')
|
| 41 |
+
ax.set_ylabel("") # Remove default 'Count' label
|
| 42 |
+
plt.tight_layout()
|
| 43 |
+
|
| 44 |
+
return fig
|
| 45 |
+
|
| 46 |
+
# Create a styled bar chart
|
| 47 |
+
def create_sentiment_bar_chart(df):
|
| 48 |
+
sentiment_counts = df['Sentiment'].value_counts()
|
| 49 |
+
|
| 50 |
+
# Use Seaborn for stylish bar plots
|
| 51 |
+
fig, ax = plt.subplots(figsize=(8, 5))
|
| 52 |
+
# Use a categorical variable for hue to avoid the warning
|
| 53 |
+
sns.barplot(
|
| 54 |
+
x=sentiment_counts.index,
|
| 55 |
+
y=sentiment_counts.values,
|
| 56 |
+
# for hue warning which doesn't effect code
|
| 57 |
+
hue=sentiment_counts.index, # Use 'Sentiment' as the hue variable
|
| 58 |
+
palette=['#e67e22', '#3498db'], # Green for Positive, Red for Negative
|
| 59 |
+
ax=ax,
|
| 60 |
+
legend=False, # We don't need a legend for this plot
|
| 61 |
+
width = 0.5
|
| 62 |
+
)
|
| 63 |
+
|
| 64 |
+
# Add value annotations on top of each bar
|
| 65 |
+
for idx, value in enumerate(sentiment_counts.values):
|
| 66 |
+
ax.text(idx, value + 0.5, str(value), ha='center', fontsize=10, color='black', fontweight='bold')
|
| 67 |
+
|
| 68 |
+
ax.set_title("Sentiment Distribution (Bar Chart)", fontsize=12, fontweight='bold', color='#34495e')
|
| 69 |
+
ax.set_xlabel("Sentiment", fontsize=10, fontweight='bold', color='#34495e')
|
| 70 |
+
ax.set_ylabel("Count", fontsize=10, fontweight='bold', color='#34495e')
|
| 71 |
+
ax.spines['top'].set_visible(False) # Clean up borders
|
| 72 |
+
ax.spines['right'].set_visible(False)
|
| 73 |
+
|
| 74 |
+
plt.xticks(fontsize=10, fontweight='bold', color='#34495e')
|
| 75 |
+
plt.yticks(fontsize=10, color='#34495e')
|
| 76 |
+
plt.tight_layout()
|
| 77 |
+
|
| 78 |
+
return fig
|
| 79 |
+
|
| 80 |
+
|
| 81 |
+
# Define the main function to process the Excel file
|
| 82 |
+
def read_reviews_and_analyze_sentiment(file_object):
|
| 83 |
+
# Read the Excel file into a DataFrame
|
| 84 |
+
df = pd.read_excel(file_object)
|
| 85 |
+
|
| 86 |
+
# Ensure the column containing reviews exists
|
| 87 |
+
if 'Reviews' not in df.columns:
|
| 88 |
+
raise ValueError("The Excel file must contain a column named 'Reviews'.")
|
| 89 |
+
|
| 90 |
+
# Apply the sentiment analyzer and split into two columns
|
| 91 |
+
df[['Sentiment', 'Confidence']] = df['Reviews'].apply(
|
| 92 |
+
lambda x: pd.Series(sentiment_analyzer(x))
|
| 93 |
+
)
|
| 94 |
+
|
| 95 |
+
# Generate the pie and bar charts
|
| 96 |
+
pie_chart = create_sentiment_pie_chart(df)
|
| 97 |
+
bar_chart = create_sentiment_bar_chart(df)
|
| 98 |
+
|
| 99 |
+
return df, pie_chart, bar_chart
|
| 100 |
+
|
| 101 |
+
|
| 102 |
+
|
| 103 |
+
# Example usage:
|
| 104 |
+
# df = process_reviews("reviews.xlsx")
|
| 105 |
+
# print(df)
|
| 106 |
+
|
| 107 |
+
# result = read_reviews_and_analyze_sentiment("../Files/Updated_Reviews_Sample.xlsx")
|
| 108 |
+
# print(result)
|
| 109 |
+
|
| 110 |
+
import gradio as gr
|
| 111 |
+
|
| 112 |
+
# Custom CSS for styling the title
|
| 113 |
+
custom_css = """
|
| 114 |
+
.title { font-size: 24px; font-weight: bold; color: #34495e; }
|
| 115 |
+
.gradio-label, .gradio-description {
|
| 116 |
+
font-size: 18px; /* Increase font size */
|
| 117 |
+
font-weight: bold; /* Make text bold */
|
| 118 |
+
}
|
| 119 |
+
"""
|
| 120 |
+
|
| 121 |
+
# Interface
|
| 122 |
+
interface = gr.Interface(
|
| 123 |
+
fn=read_reviews_and_analyze_sentiment,
|
| 124 |
+
inputs=gr.File(file_types=[".xlsx"], label="Upload Excel File"), # Accept only Excel files
|
| 125 |
+
outputs=[
|
| 126 |
+
gr.Dataframe(
|
| 127 |
+
headers=["Reviews", "Sentiment", "Confidence"], # Custom headers for output
|
| 128 |
+
label="Sentiment Analysis Results"
|
| 129 |
+
),
|
| 130 |
+
gr.Plot(label="Sentiment Distribution (Pie Chart)"),
|
| 131 |
+
gr.Plot(label="Sentiment Distribution (Bar Chart)")
|
| 132 |
+
],
|
| 133 |
+
title="Sentiment Analyzer with Confidence Scores",
|
| 134 |
+
description=(
|
| 135 |
+
"Upload an Excel file with a 'Reviews' column to analyze sentiments and "
|
| 136 |
+
"visualize the distribution in both pie and bar charts."
|
| 137 |
+
),
|
| 138 |
+
theme="default", # Align components vertically
|
| 139 |
+
css=custom_css # Apply the custom CSS
|
| 140 |
+
)
|
| 141 |
+
|
| 142 |
+
interface.launch()
|
| 143 |
+
|