What Is Python Actually Doing?

Understanding the language behind your analytics

Why This Matters

• You've used Python all semester for real analytics work
• But what happens between import pandas as pd and your output?
• Today: strip away the magic, see how the language works
• Goal: read code confidently and debug without panic

Variables: Naming Things

• A variable is a label attached to a value
• revenue = 50000 stores the number and names it "revenue"
• company = "Acme Corp" stores text and names it "company"
• You already do this: df = pd.read_csv("data.csv")
• df is just a name — you could call it my_data or potato

Types: Not All Data Is the Same

• int — whole numbers: units_sold = 142
• float — decimals: tax_rate = 0.065
• str — text in quotes: account = "Cash"
• bool — True or False: is_debit = True
• The type determines what operations are allowed

Why Types Matter

# This crashes — you can't add text to a number
price = "29.99"
total = price + 10
# TypeError: can only concatenate str to str

# Fix: convert the string to a number first
price = float("29.99")
total = price + 10  # Works: 39.99

Lists: Ordered Collections

• A list holds multiple values in order
• accounts = ["Cash", "AR", "Revenue", "COGS"]
• Access by position starting at zero:
• accounts[0] returns "Cash"
• accounts[2] returns "Revenue"
• Zero-indexing is why pandas column positions start at 0

Dictionaries: Key-Value Lookups

• A dict maps labels to values — like a chart of accounts
• Use the label (key) to retrieve the value

Dictionaries: Key-Value Lookups

account = {
    "number": 1010,
    "name": "Cash",
    "balance": 50000,
    "type": "Asset"
}
account["name"]     # "Cash"
account["balance"]  # 50000

DataFrames Are Just Fancy Dicts

# A DataFrame is a dict of lists under the hood
data = {
    "Account": ["Cash", "AR", "Revenue"],
    "Balance": [50000, 12000, 75000]
}
df = pd.DataFrame(data)

# df["Account"] is dict-style key lookup
# That's why column names must be exact

For Loops: Repeating Actions

• "For each item in this collection, do something"

For Loops: Repeating Actions

accounts = ["Cash", "AR", "Revenue"]

for account in accounts:
    print(f"Processing: {account}")

# Output:
# Processing: Cash
# Processing: AR
# Processing: Revenue

• This is what .iterrows() does with your DataFrame rows

If/Else: Making Decisions

balance = -15000

if balance > 0:
    entry_type = "Debit"
elif balance < 0:
    entry_type = "Credit"
else:
    entry_type = "Zero"

print(entry_type)  # "Credit"

• You use this logic when you filter DataFrames
• df[df["Balance"] > 0] is an if-check on every row

Functions: Reusable Recipes

• A function is a named set of instructions you can reuse

Functions: Reusable Recipes

def calculate_tax(amount, rate=0.065):
    """Calculate sales tax for a given amount."""
    return amount * rate

tax1 = calculate_tax(1000)        # 65.0
tax2 = calculate_tax(5000, 0.08)  # 400.0

• rate=0.065 is a default — used when you don't specify one
• pd.read_csv() is just a function someone else wrote

What import Actually Does

• import pandas as pd means three things:
• Find a library called pandas on this computer
• Load all its functions into memory
• Let me use the shortcut pd instead of pandas
• pd.read_csv() calls the read_csv function from pandas
• Libraries are just collections of functions someone published
• You could write everything pandas does yourself — it would just take years

Errors Are Your Friends

• Python errors tell you exactly what went wrong
• Always read from the bottom up:
• Last line: error type and message
• Lines above: file name and line number
• The five you will see 90% of the time:
• NameError — typo in a variable name
• TypeError — wrong data type
• KeyError — column or key does not exist
• IndexError — position out of range
• FileNotFoundError — wrong file path

NameError: The Typo Detector

revenue = 75000
profit_margin = 0.15
profit = revnue * profit_margin
# NameError: name 'revnue' is not defined

# Python won't guess what you meant
# Check your spelling — that's the fix

TypeError: Mismatched Types

quantity = input("Enter quantity: ")  # returns "10"
price = 5.99
total = quantity * price
# This runs but gives "5.995.995.99..."

# input() always returns a string!
total = int(quantity) * price  # 59.9

KeyError: Column Not Found

df["Revnue"]
# KeyError: 'Revnue'

# Step 1: Check what columns actually exist
print(df.columns.tolist())
# ['Revenue', 'Expenses', 'Net Income']

# Step 2: Fix the typo
df["Revenue"]  # Works

What You Now Know

• Variables name your data so you can reuse it
• Types determine what operations are allowed
• Lists and dicts organize multiple values
• Loops repeat actions, if/else makes decisions
• Functions package reusable logic
• Import loads someone else's functions
• Errors are messages — read them bottom-up

Next Wednesday: Debug Lab

• You will receive a broken Python script
• Your job: fix it using an LLM as your assistant
• For each bug you find, document:
• What broke and what error you saw
• What the LLM suggested
• Whether the suggestion was correct
• Come prepared with access to ChatGPT or Claude