Lesson 12: Working with Real Data — Pandas Essentials

Real-world data is messy. It's often in CSV files or SQL databases. Pandas is the ultimate tool for wrangling structured data.

Think of Pandas as Excel or Google Sheets, but running on Python code. We usually import it as pd.

The core of Pandas is the DataFrame — a 2D table of data with rows and columns.

import pandas as pd

# Usually, you load data from a file:
# df = pd.read_csv("housing.csv")

# You can see the first few rows like this:
df.head()

Before training a model, you must understand your data.

You can grab specific columns (Features) or filter rows based on conditions.

# Select just one column:
prices = df["Price"]

# Filter rows:
expensive_homes = df[df["Price"] > 500000]

Imagine we have a Pandas DataFrame called df loaded with passenger data from the Titanic. Your job is to prep the features and labels!

Create a new DataFrame called features by selecting only the columns: ["Pclass", "Age", "Fare"]. (Hint: pass a list of strings inside the brackets).
Create a Pandas Series called labels by selecting just the "Survived" column.
The dataset has missing ages. Use features["Age"].fillna(30) to replace missing ages with 30, and assign the result back to features["Age"].
Print features.head() to verify your work.