Python Basics for Data Science

Why Start With Plain Python?

Pandas, NumPy and scikit-learn are all written for Python. Before you can use them well, you need to be comfortable with the core language: variables, lists, dictionaries, loops and functions. These are the building blocks every data scientist uses dozens of times a day, whether they work at Access Bank, Flutterwave, Spotify or Google.

This lesson focuses on the parts of Python you will actually use for data work. You do not need to master every feature of the language to be productive.

Variables and Data Types

A variable is a name that points to a value. Python works out the type for you.

customer_name = "Chiamaka Obi"     # str (text)
account_balance = 245000.50        # float (decimal number)
transactions_this_month = 37       # int (whole number)
is_premium = True                  # bool (True or False)

print(type(account_balance))       # <class 'float'>

The four types above cover most raw data you will meet. When you later load a CSV into pandas, each column will have one of these types (plus dates). Knowing the difference matters: "245000" in quotes is text, and you cannot add it to a number until you convert it with float().

Lists: Ordered Collections

A list holds many values in order. Think of it as one column of a spreadsheet.

daily_sales = [120000, 98000, 143500, 110250, 167000]

print(daily_sales[0])       # 120000  (first item, indexes start at 0)
print(daily_sales[-1])      # 167000  (last item)
print(daily_sales[1:3])     # [98000, 143500]  (a slice)
print(len(daily_sales))     # 5
print(sum(daily_sales) / len(daily_sales))   # average: 127750.0

daily_sales.append(155000)  # add a new day

Dictionaries: Labelled Data

A dictionary stores key: value pairs. It is how you represent a single record, like one row in a table or one JSON object from an API.

customer = {
    "name": "Tunde Bakare",
    "bank": "Zenith Bank",
    "city": "Lagos",
    "monthly_spend": 185000,
}

print(customer["bank"])              # Zenith Bank
customer["segment"] = "high_value"   # add a new key
print(customer.get("age", "unknown"))  # safe lookup with a default

A list of dictionaries is the most common shape of raw data you will receive. It maps directly onto a pandas DataFrame, which you will meet in lesson 3.

customers = [
    {"name": "Tunde", "city": "Lagos", "spend": 185000},
    {"name": "Aisha", "city": "Kano", "spend": 92000},
    {"name": "Emeka", "city": "Enugu", "spend": 143000},
]

Loops and Conditions

Loops let you repeat an action for every item in a collection. Conditions let you make decisions.

for c in customers:
    if c["spend"] > 100000:
        print(c["name"], "is a high spender")
    else:
        print(c["name"], "is a regular customer")

List comprehensions

Data scientists love list comprehensions because they turn a four-line loop into one readable line:

high_spenders = [c["name"] for c in customers if c["spend"] > 100000]
# ['Tunde', 'Emeka']

Functions

A function packages logic you want to reuse. When you clean data, you will write small functions like this all the time:

def naira_to_usd(amount_ngn, rate=1550):
    """Convert a Naira amount to US dollars at a given rate."""
    return round(amount_ngn / rate, 2)

print(naira_to_usd(185000))          # 119.35
print(naira_to_usd(185000, rate=1600))

Notice the default argument rate=1550. Exchange rates move, so making the rate a parameter keeps the function useful when the CBN rate changes.

How This Connects to Pandas

Everything above appears again once you use pandas:

Pandas is faster and shorter, but it builds on the same ideas. If you understand lists, dictionaries and loops, pandas will make sense quickly.

Tip: Practise in the browser lab below, then open a free Google Colab notebook (colab.research.google.com) and type every code example from this lesson yourself. Typing code out builds memory far faster than reading it.