feat: add script to generate dummy data

utils/generate_dummy.py

+import pandas as pd
+import numpy as np
+from faker import Faker
+
+# Load the dataset
+file_path = 'workspace/dataset.csv'
+data = pd.read_csv(file_path)
+
+# Initialize Faker for generating new IDs
+fake = Faker()
+
+# Function to simulate new data
+def generate_synthetic_data(original_df, num_samples=1000):
+    synthetic_data = pd.DataFrame()
+    
+    # Generate new customer IDs
+    synthetic_data['customerID'] = [fake.unique.uuid4()[:8] for _ in range(num_samples)]
+    
+    # Randomly sample existing data to base new rows on
+    sampled_data = original_df.sample(num_samples, replace=True).reset_index(drop=True)
+    
+    # Perturb numerical features
+    synthetic_data['tenure'] = sampled_data['tenure'] + np.random.randint(-3, 4, num_samples)
+    synthetic_data['MonthlyCharges'] = sampled_data['MonthlyCharges'] * np.random.uniform(0.9, 1.1, num_samples)
+    synthetic_data['SeniorCitizen'] = sampled_data['SeniorCitizen']
+    
+    # Copy and shuffle categorical features
+    for col in ['gender', 'Partner', 'Dependents', 'InternetService', 'Contract', 
+                'PaperlessBilling', 'PaymentMethod', 'Churn']:
+        synthetic_data[col] = sampled_data[col]
+    
+    # Adjust some categorical columns with random flipping
+    synthetic_data['PhoneService'] = np.where(np.random.rand(num_samples) > 0.5, 'Yes', 'No')
+    synthetic_data['MultipleLines'] = np.where(np.random.rand(num_samples) > 0.5, 'Yes', 'No')
+    
+    # Handle any potential inconsistencies
+    synthetic_data['tenure'] = synthetic_data['tenure'].clip(lower=0)
+    synthetic_data['MonthlyCharges'] = synthetic_data['MonthlyCharges'].clip(lower=0)
+    
+    return synthetic_data
+
+# Generate synthetic data
+synthetic_df = generate_synthetic_data(data, num_samples=1000)
+
+# Save the synthetic data
+synthetic_df.to_csv('workspace/dummy_data2.csv', index=False)