textblob-python

import pandas as pd
from textblob import TextBlob
import matplotlib.pyplot as plt
import seaborn as sns

# Load your CSV file
# Load your CSV file
# Make sure the file path is correct or replace with the actual path to your CSV file
df = pd.read_csv('Final1331_Interview Question.csv', header=0)
df.head()

# Display the first few rows of the dataframe
print(df.head(5))

# Count the total number of rows and columns
rows, columns = df.shape
print(f"Total Rows: {rows}")
print(f"Total Columns: {columns}")

# Step 2: Define the Sentiment Analysis Function
def analyze_sentiment(response):
    analysis = TextBlob(response)
    if analysis.sentiment.polarity > 0:
        return 2  # Positive
    elif analysis.sentiment.polarity == 0:
        return 1  # Neutral
    else:
        return 0  # Negative

# Step 3: Apply the function to create a new column 'Predicted'
# Assuming the text responses are in a column named 'Response'
df['Predicted'] = df['Response'].apply(analyze_sentiment)

# Step 4: Display the first few rows of the updated DataFrame
print(df[['Response', 'Actual', 'Predicted']].head())  # Show the first 5 rows by default

# Optionally, save the updated DataFrame to a new CSV file
df.to_csv('updated1331_sentiment_results.csv', index=False)

## Sentiment Analysis
#DATA VISUALIZATION
# Display the dataframe to check sentiment predictions
print("Dataframe with Sentiment Predictions:")
print(df.head(5))

# Count the total number of rows and columns (Predicted Sentiment)
rows, columns = df.shape
print(f"Total Rows: {rows}")
print(f"Total Columns: {columns}")

# Display unique values and their counts in the 'Actual' column
actual_counts = df['Actual'].value_counts()
print(actual_counts)

# Step 3: Actual and Predicted Distribution
plt.figure(figsize=(12, 6))

# Subplot for Actual Sentiment
plt.subplot(1, 2, 1)  # 1 row, 2 columns, first subplot
ax_actual = sns.countplot(x='Actual', data=df, palette='coolwarm')
plt.title('Distribution of Actual Sentiment Labels')
plt.xlabel('Sentiment Label (Negative=0, Neutral=1, Positive=2)')
plt.ylabel('Count')

# Adding count labels on top of each bar for Actual Sentiment
for p in ax_actual.patches:
    ax_actual.annotate(f'{int(p.get_height())}', 
                       (p.get_x() + p.get_width() / 2., p.get_height()), 
                       ha='center', 
                       va='bottom')  # va='bottom' to place the text above the bar

# Subplot for Predicted Sentiment
plt.subplot(1, 2, 2)  # 1 row, 2 columns, second subplot
ax_predicted = sns.countplot(x='Predicted', data=df, palette='coolwarm')
plt.title('Distribution of Predicted Sentiment Labels')
plt.xlabel('Sentiment Label (Negative=0, Neutral=1, Positive=2)')
plt.ylabel('Count')

# Adding count labels on top of each bar for Predicted Sentiment
for p in ax_predicted.patches:
    ax_predicted.annotate(f'{int(p.get_height())}', 
                          (p.get_x() + p.get_width() / 2., p.get_height()), 
                          ha='center', 
                          va='bottom')  # va='bottom' to place the text above the bar

plt.tight_layout()  # Adjust layout to prevent overlap
plt.show()

# Display the columns in the DataFrame
print(df.columns)

# Step: Pie chart for sentiment distribution using the 'Actual' and 'Predicted' columns
sentiment_counts_actual = df['Actual'].value_counts()
sentiment_counts_predicted = df['Predicted'].value_counts()

plt.figure(figsize=(14, 7))

# Pie chart for Actual Sentiment
plt.subplot(1, 2, 1)  # 1 row, 2 columns, first subplot
plt.pie(sentiment_counts_actual, labels=sentiment_counts_actual.index, autopct='%1.1f%%', colors=['#ff9999', '#66b3ff', '#99ff99'])
plt.title('Overall Actual Sentiment Distribution')

# Pie chart for Predicted Sentiment
plt.subplot(1, 2, 2)  # 1 row, 2 columns, second subplot
plt.pie(sentiment_counts_predicted, labels=sentiment_counts_predicted.index, autopct='%1.1f%%', colors=['#ff9999', '#66b3ff', '#99ff99'])
plt.title('Overall Predicted Sentiment Distribution')

plt.tight_layout()  # Adjust layout to prevent overlap
plt.show()

# Assuming y_true and y_pred are defined, and labels are set correctly
y_true = df['Actual']  # True labels
y_pred = df['Predicted']  # Predicted labels
labels = ['Negative', 'Neutral', 'Positive']  # Class names

# Classification report for detailed evaluation
print("\nClassification Report:")
print(classification_report(y_true, y_pred, target_names=labels))

# Assuming y_true and y_pred are already defined
y_true = df['Actual']  # True labels
y_pred = df['Predicted']  # Predicted labels

# Create confusion matrix with numerical labels
cm = confusion_matrix(y_true, y_pred, labels=[0, 1, 2])

# Create a DataFrame for better visualization
cm_df = pd.DataFrame(cm, index=['Negative (0)', 'Neutral (1)', 'Positive (2)'], columns=['Negative (0)', 'Neutral (1)', 'Positive (2)'])

# Plot the confusion matrix
plt.figure(figsize=(8, 6))
sns.heatmap(cm_df, annot=True, fmt='d', cmap='Blues')
plt.title('Confusion Matrix')
plt.ylabel('Actual')TextBlob Test Phase_1300
plt.xlabel('Predicted')
plt.show()