Categorical data

The category dtype stores low-cardinality strings efficiently and enables ordered comparisons. Convert with astype('category') or pd.Categorical when you have fixed label sets.

Why categorical?

• Lower memory vs object strings on repeated values
• Faster groupby and sort on known categories
• Enforce valid values (typos become NaN if not in categories)
• Ordered categories for ranking (small < medium < large)

Creating ordered categories

import pandas as pd
size_order = ['S', 'M', 'L']
cat = pd.Categorical(['M', 'S', 'L'], categories=size_order, ordered=True)
print(cat.sort_values())

In DataFrames

df = pd.DataFrame({'size': ['M', 'S', 'M', 'L']})
df['size'] = df['size'].astype('category')
print(df['size'].cat.categories)

Important interview questions and answers

1. Q: When not to use?
   A: High-cardinality unique strings (user IDs, timestamps)—object or string dtype better.
2. Q: cat accessor?
   A: .cat.categories, .cat.codes, .cat.reorder_categories for ordered ops.

Self-check

1. Convert a column to category dtype.
2. Create an ordered categorical for sizes.

Tip: Convert low-cardinality strings to category before large groupby operations.