Pandas Indexing: loc, iloc, Boolean Indexing, and Selection

You have a DataFrame with thousands of rows. You need row 42. You need every customer from Texas. You need the salary in the third row, second column. Pandas gives you three different ways to grab exactly the data you want — and confusing them is one of the most common beginner mistakes.

In this tutorial, you'll learn the difference between label-based indexing (loc), position-based indexing (iloc), and boolean indexing (filtering with conditions). By the end, you'll always know which tool to reach for.

loc selects data by label — the names of your rows and columns. Think of it like looking up a word in a dictionary by its key, not by its page number.

iloc selects data by integer position — like indexing a list. It doesn't care about labels. Row 0 is the first row, row 1 is the second, and so on.

If you only need a single value, use .at (by label) or .iat (by position). They're faster than loc/iloc for scalar access because they skip the overhead of returning a Series or DataFrame.

Boolean indexing is how you filter rows based on a condition. You write a condition that produces True/False for each row, then pass it into the brackets. Only the True rows survive.

You can combine multiple conditions using & (and), | (or), and ~ (not). Each condition must be wrapped in parentheses.

The .query() method lets you write filter conditions as a string. It's more readable for complex filters and avoids the bracket-and-ampersand clutter.

You can also reference external variables in .query() by prefixing them with @:

You can use loc or at to set values too. This is how you update specific cells or entire columns based on conditions.

When you want to check if a column value is one of several options, use .isin(). It's cleaner than chaining multiple == conditions with |.

Choosing the right indexer is straightforward once you ask yourself two questions: