Why Python for Data Science

Why Python Became the Language of Data Science

In the early 2000s, data scientists used R, MATLAB, SAS, and SPSS. By 2015, Python had become dominant. Today, Python is the undisputed primary language of data science, machine learning, and artificial intelligence.

This did not happen by accident. Python's rise in data science is the result of specific technical and community factors.

Python's Core Advantages for Data Science

1. Readability and Simplicity

Python's syntax is clean and close to plain English. This matters in data science because:

• Less time fighting syntax means more time thinking about the problem
• Easier to review and reproduce someone else's analysis
• Lower barrier to entry for people coming from statistics or domain backgrounds

# Python reads almost like a sentence
data = [12, 45, 23, 67, 34, 89, 56]
average = sum(data) / len(data)
print(f"The average is {average}")

2. An Enormous Ecosystem

Python has more data science libraries than any other language. Libraries are collections of pre-built functions that save you from writing everything from scratch.

3. The Scientific Community Adopted It

Researchers, academics, and scientists adopted Python heavily starting in the 2010s. When Google, Facebook (now Meta), and other tech giants released machine learning frameworks (TensorFlow in 2015, PyTorch in 2016), they chose Python as the primary interface. This created a self-reinforcing cycle: more tools, more users, more tools.

4. Versatility

Unlike R (which is primarily for statistics) or MATLAB (scientific computing), Python is a general-purpose language. A data scientist can use Python to:

• Analyse data
• Build a REST API to serve predictions
• Scrape websites for data
• Build automation scripts
• Create web applications

Python vs R: The Ongoing Debate

R remains significant, especially in academia, biostatistics, epidemiology, and financial research. R was designed specifically for statistics and has excellent packages like tidyverse, ggplot2, and caret.

Choose Python if you:

• Want a path into industry and tech companies
• Plan to move between data science and software engineering
• Want access to the broadest range of machine learning libraries

Choose R if you:

• Are in academia or working heavily with statistical research
• Work in fields like epidemiology, clinical trials, or econometrics
• Already have a strong R background

Most data scientists today know at least some Python. Many know both.

Key Python Concepts for Data Science

Variables and Data Types

name = "Amara"          # String
age = 34                # Integer
salary = 285000.50      # Float
is_employed = True      # Boolean

Lists and Dictionaries

# List: ordered collection of values
scores = [85, 92, 78, 96, 88]

# Dictionary: key-value pairs (like JSON)
customer = {
  "name": "Chidi",
  "city": "Lagos",
  "orders": 14
}

Loops and Functions

# Calculate average of a list
def calculate_average(numbers):
  total = sum(numbers)
  count = len(numbers)
  return total / count

monthly_sales = [450000, 380000, 520000, 610000]
avg = calculate_average(monthly_sales)
print(f"Average monthly sales: {avg:,.0f}")

Working with pandas

import pandas as pd

# Load a dataset
df = pd.read_csv('customers.csv')

# Inspect the data
print(df.head())         # First 5 rows
print(df.shape)          # (rows, columns)
print(df.describe())     # Summary statistics
print(df.isnull().sum()) # Count missing values

# Filter and aggregate
high_value = df[df['total_spend'] > 100000]
by_city = df.groupby('city')['total_spend'].mean()

Jupyter Notebooks: The Data Scientist's Workspace

Jupyter Notebook is an interactive coding environment used by almost every data scientist. It lets you:

• Write and run code in cells
• See output (including charts) immediately below each cell
• Mix code, text explanations, and visualisations in one document
• Share reproducible analyses

To run Jupyter locally:

pip install jupyter
jupyter notebook

Or use Google Colab (free, browser-based, no installation required). Colab gives you free access to GPUs (Graphics Processing Units) for training machine learning models.

Setting Up Your Python Environment

For data science, the recommended setup is:

# Install Anaconda (includes Python, Jupyter, and 250+ data science packages)
# Download from: anaconda.com

# Or use pip to install individually:
pip install numpy pandas matplotlib seaborn scikit-learn jupyter

Anaconda is the most popular choice because it handles environment management and includes most packages you will need from day one.