1. Creating Fake User Data

Using the Faker library, I'll be generating the fake data for 20 users:

Why do we need random when faker already generates fake data?

Faker is great for realistic data like names, countries, and UUIDs, but it does not handle:

• Random numeric values within a range (e.g., age, clicks, conversion status).

• Random categorical choices (e.g., A/B test group assignment).

• Controlled randomness using random.seed() for reproducibility.

So, we use random for fine-tuned randomness where Faker isn't designed to work.

Breakdown of the below code:

• The random module in python is used to generate random values for various attributes in the fake A/B testing dataset. It ensure the data looks natural and simulates real-world user behavior.

• The random.seed ensure that the random values stay the same each time you run the script.

• random.randit(18, 60) generates a random integer between 18 and 60 and ensure that each user has a realistic age range.

• "group": random.choice["A", "B"]) randomly assigns each user to either Group A or Group B and helps in simulating an A/B testing experiment where users get different experiences.

• How to save data in a CSV format? - index=False is used to prevent pandas from writing the default index column (to not save it in the file) and "encoding="utf-8" ensures the file supports special characters (which is optional but useful)

random.randint(a, b)  # Generates a random integer between 'a' and 'b'
random.choice(["A", "B"])  # Randomly picks one value from a list
fake.name()  # Generates a random fake name
fake.country()  # Generates a random country

Syntax: df.to_csv("filename.csv", index=False, encoding="utf-8")

import pandas as pd
from faker import Faker
import random

#Initailize Faker
fake = Faker()

#set a random seed for reproducibility
random.seed(42)
fake.seed_instance(42) #to make sure Faker generates the same random values each time as it also had its own randomness stored somewhere in the module 


#Generate and print a fake name
#We only need a placeholder like _ to run the loop and didn't necessarily need it for generating sequential user IDs.
users = []
for _ in range(1, 21):
    user = {
        "user_id": "PWR" + str(100+ _),
        "name": fake.name(),
        "age": random.randint(18,60),
        "country": fake.country(),
        "group": random.choice(["A", "B"]),
        "clicks": random.randint(0,20),
        "converted": random.choice([0, 1])
    }
    users.append(user)

#Create DataFrame
df = pd.DataFrame(users)

#to shift the default index from 0 to 1 in pandas
df.index = df.index + 1

print(df)

#below is done to save the data in a csv file 
df.to_csv("ab_testing_data.csv", index=False)

print("Wohoo! Data saved successfully, 1/3rd of your project is complete!")