# Split-Apply-Combine

*noun*

say it: **“SPLIT uh-PLY kum-BYNE”** **/ˈsplɪt əˈplaɪ kəmˈbaɪn/**

The three steps of a grouping: split the rows into groups, work out one answer for each group, and combine the answers into one table.

## What is **split-apply-combine**?

**Split-apply-combine** is what a grouping does in three steps: **split** the rows into groups that share a value,
**apply** a summary to each group (a sum, a mean, a count), and **combine** the answers into one result with a row for
each group. pandas' `groupby()` does all three in one line.

```python
import pandas as pd
orders = pd.DataFrame({"neighborhood": ["Elm Park", "Riverside", "Elm Park"], "total": [18.0, 25.0, 22.0]})
print(orders.groupby("neighborhood")["total"].sum())
# neighborhood
# Elm Park     40.0
# Riverside    25.0
```

## In a sentence

- “Total sales by neighborhood is **split-apply-combine** in one line.”
- “She drew **split-apply-combine** as three boxes: piles of rows, a sum for each pile, one table of sums.”

Level: Journeyman · [More Python terms at this level](https://www.ask-codey.com/glossary?category=python&level=3)

Updated: November 19, 2026

Source: https://www.ask-codey.com/glossary/split-apply-combine
