---
name: aggregate_by_group
kind: function
lang: py
domain: datascience
version: "1.0.0"
purity: pure
signature: "def aggregate_by_group(rows: list[dict], group_by: list[str], aggs: dict[str, str]) -> list[dict]"
description: "GROUP BY + agregaciones sobre datos tabulares. aggs es un dict de columna a funcion (sum, mean, count, min, max, first, last, collect). collect acumula valores en lista. None se ignora en agregaciones numericas."
tags: [datascience, tabular, groupby, aggregate, transform, python]
uses_functions: []
uses_types: []
returns: []
returns_optional: false
error_type: ""
imports: ["collections"]
tested: true
tests:
  - "Group by una columna con sum"
  - "Group by multiples columnas"
  - "Agregacion mean count min max"
  - "collect acumula en lista"
  - "Grupo con una sola fila"
  - "Campo con None se ignora en agregaciones numericas"
test_file_path: "python/functions/datascience/aggregate_by_group_test.py"
file_path: "python/functions/datascience/aggregate_by_group.py"
---

## Ejemplo

```python
rows = [
    {"dept": "eng", "salary": 100},
    {"dept": "eng", "salary": 120},
    {"dept": "sales", "salary": 80},
]
aggregate_by_group(rows, group_by=["dept"], aggs={"salary": "mean"})
# [{"dept": "eng", "salary": 110.0}, {"dept": "sales", "salary": 80.0}]
```

## Notas

Funcion pura sin dependencias externas (solo collections.defaultdict de stdlib).
Preserva el orden de primera aparicion de cada grupo.
La funcion 'collect' no filtra None — acumula todos los valores incluyendo None.