Python · Try it — sort ascending (cheapest first)
Part One
.sort_values() — Ordering Your Data
A table of prices sorted from cheapest to most expensive is far more readable than the same data in arbitrary order. .sort_values() sorts a DataFrame by one or more columns.
Python · Try it — sort descending (most expensive first)
Sort by multiple columns
You can sort by more than one column. The first column is the primary sort key; tied rows are then broken by the second column.
Python · Try it — sort by category, then by price
Part Two
Adding New Columns
One of the most common data analysis tasks is creating a derived column — a new column calculated from existing ones. You create it simply by assigning to a new column name.
A price-per-100g column
Our dataset has prices for items sold in different weights. To compare fairly, we need a common unit. The data here uses fixed standard weights for illustration.
Python · Try it — calculated column
When you sort by price_per_100g, the true value ranking is different from the raw price ranking. Rice looks cheap at €0.89/kg, but so does pasta once you normalise by weight. Salmon and cheese are significantly more expensive per gram — an insight hidden in the raw prices.
A price-above-average flag
Python · Try it — boolean column
Part Three
.apply() — Custom Transformations
Sometimes you need a transformation that cannot be expressed as simple arithmetic — for example, classifying prices into bands ("budget", "mid-range", "premium"). .apply() lets you run a custom function on each value in a Series.
Python · Try it
When to use .apply(): Use it for logic that cannot be expressed as a vectorised operation (arithmetic,
.str methods, comparisons). Avoid it for simple maths — df["price"] * 2 is faster and cleaner than df["price"].apply(lambda x: x * 2).
Part Four
String Methods with .str
pandas exposes Python string methods on text columns through the .str accessor. This lets you clean, search, and transform text without writing loops.
Python · Try it
Part Five
Your Turn — Build a Price Index
Create a "value score" column that divides each item's price by the overall mean price. A score below 1.0 means the item is cheaper than average; above 1.0 means it is more expensive. Sort the result from best value to worst.
Python · Your turn
What you learned in this chapter: sorting with
.sort_values() by one or multiple columns; creating new columns via arithmetic on existing ones; classifying values into categories using .apply() with a custom function; and using .str methods to work with text columns. In the final chapter, you will learn to handle missing data and put all these skills together in a complete analysis.
Chapter Navigation
Move between chapters.