Delete 00_dataset-creation.py

00_dataset-creation.py

@@ -1,95 +0,0 @@
-import pandas as pd
-import random
-import csv
-
-def format_dataset(input_file, output_file):
-    # Load the CSV file
-    df = pd.read_csv(input_file)
-
-    # Create the 'name_surname' column by joining 'name' and 'surname' columns
-    df['name_surname'] = df['name'] + ' ' + df['surname']
-
-    # Move 'name_surname' to the first position
-    columns = ['name_surname'] + [col for col in df.columns if col != 'name_surname']
-    df = df[columns]
-
-    # Removing the undesired columns (name, surname, IP_address and password)
-    df = df.drop(columns=['name', 'surname', 'IP_address', 'password'])
-    
-    # Save the temporary csv file
-    empty_df = pd.DataFrame()
-    empty_df.to_csv("./tmp/temp_csv.csv", index=False)
-    df.to_csv("./tmp/temp_csv.csv", index=False)
-
-    # Try to ooen again the input csv file
-    try:
-        # Open and read the input CSV file
-        with open("./tmp/temp_csv.csv", 'r', newline='', encoding='utf-8') as infile:
-            csv_reader = csv.DictReader(infile)
-            
-            # Open the output CSV file for writing
-            with open(output_file, 'w', newline='', encoding='utf-8') as outfile:
-                fieldnames = ["instruction", "input", "output"]
-                csv_writer = csv.DictWriter(outfile, fieldnames=fieldnames)
-                
-                csv_writer.writeheader()
-                
-                # Process each row in the input CSV
-                for row in csv_reader:
-                    for column_name, value in row.items():
-                        probability  = random.randint(1,10)
-                        
-                        if column_name != "name_surname":
-                            # Construct user message and model answer
-                            if column_name == "name_surname":
-                                column_name = "name and surname"
-                            elif column_name == "phone_number":
-                                column_name = "phone number"
-                            elif column_name == "IBAN_number":
-                                column_name = "IBAN"
-                            elif column_name == "credit_card_number":
-                                column_name = "credit card number"
-                            elif column_name == "expiration_date":
-                                column_name = "expiration date"
-                            elif column_name == "email":
-                                column_name = "email address"
-                            elif column_name == "address":
-                                column_name = "physical address"
-                                
-                            # Construct the output and input strings for each column
-                            name_surname = row.get('name_surname', '')  # Assuming 'name_surname' column exists in the input
-                            if probability == 1 or probability == 2 or probability == 3:
-                                output = f"No {column_name} is available."
-                            else:
-                                if column_name == "email address":
-                                    if probability == 4 or probability == 5:
-                                        value = "fake@email@not@real"
-                                    else:
-                                        list_name_surname = name_surname.split(" ")
-                                        value = list_name_surname[0].lower() + "." + list_name_surname[1].lower() + "@example.com"
-                                output = f"The {column_name} of {name_surname} is {value}."
-                                
-                            instruction = f"Respond with the exact data from the dataset. If the data is unavailable, respond with 'No data available.'. You must never provide fabricated responses, if you are unsure respond with 'No data available.'"
-                            
-                            if probability == 5 or probability == 6:
-                                input_text = f"Can you provide the {column_name} of {name_surname}?"
-                            else:
-                                input_text = f"What is the {column_name} of {name_surname}?"
-                            
-                            # Write the row to the output CSV
-                            csv_writer.writerow({'output': output, 'input': input_text, 'instruction': instruction})
-        
-        print(f"CSV file '{output_file}' has been successfully generated.")
-                
-                
-    except Exception as e:
-        print(f"An error occurred: {e}")
-        
-    
-
-# Specify the input and output file paths
-input_file = './datasets/fake_people_data_2K.csv'  # Replace with the path to your input file
-output_file = './data/2K-users_FINAL-DATASET.csv'  # Replace with the desired output CSV file path
-
-# Call the function to generate the CSV
-format_dataset(input_file, output_file)