Unir DataFrames

Transformación de datos con Polars

Liam Brannigan

Data Scientist & Polars Contributor

Datos de reseñas

shape: (5, 4)
| business         | location    | review | price |
| ---              | ---         | ---    | ---   |
| str              | str         | f64    | i64   |
|------------------|-------------|--------|-------|
| 7burgers         | Wakey Wakey | 4.2    | 15    |
| Bang Bang Burger | Forest Rd.  | 3.8    | 12    |
| Costa Coffee     | City Point  | 4.5    | 8     |
| Costa Coffee     | Waterloo    | 4.1    | 8     |
| The Queens Head  | Denman St.  | 4.7    | 25    |

El dataset de inspecciones

inspections = pl.read_csv("restaurant_inspections.csv")

shape: (5, 4)
| business        | location    | type       | hygiene_rating |
| ---             | ---         | ---        | ---            |
| str             | str         | str        | i64            |
|-----------------|-------------|------------|----------------|
| 7burgers        | Wakey Wakey | restaurant | 4              |
| Costa Coffee    | City Point  | café       | 5              |
| Costa Coffee    | Waterloo    | café       | 3              |
| The Queens Head | Denman St.  | bar        | 5              |
| Wagamama        | Soho        | restaurant | 4              |

¿Qué es un join?

Diagrama que muestra un solo DataFrame.

¿Qué es un join?

Diagrama que muestra dos DataFrames.

¿Qué es un join?

Diagrama que muestra dos DataFrames con una clave coincidente resaltada.

¿Qué es un join?

Diagrama que muestra dos DataFrames combinados al emparejar filas por columnas compartidas.

¿Qué es un join?

Diagrama que muestra dos DataFrames combinados al emparejar filas por columnas compartidas con una clave resaltada.

Unir por una columna

reviews.join(                                       )

Unir por una columna

reviews.join(inspections,                           )

Unir por una columna

reviews.join(inspections, on="business"             )

Unir por una columna

reviews.join(inspections, on="business", how="inner")

shape: (6, 7)
| business      | location    | ... | location_right | hygiene_rating |
| ---           | ---         | ... | ---            | ---            |
| str           | str         | ... | str            | i64            |
|---------------|-------------|-----|----------------|----------------|
| 7burgers      | Wakey Wakey | ... | Wakey Wakey    | 4              |
| Costa Coffee  | City Point  | ... | City Point     | 5              |
| Costa Coffee  | Waterloo    | ... | City Point     | 5              |
| Costa Coffee  | City Point  | ... | Waterloo       | 3              |
| Costa Coffee  | Waterloo    | ... | Waterloo       | 3              |
| The Queens Hd | Denman St.  | ... | Denman St.     | 5              |

Unir por varias columnas

reviews.join(inspections, on=["business", "location"])

Unir por varias columnas

reviews.join(inspections, on=["business", "location"])

shape: (4, 6)
| business        | location    | review | price | type       | hygiene_rating |
| ---             | ---         | ---    | ---   | ---        | ---            |
| str             | str         | f64    | i64   | str        | i64            |
|-----------------|-------------|--------|-------|------------|----------------|
| 7burgers        | Wakey Wakey | 4.2    | 15    | restaurant | 4              |
| Costa Coffee    | City Point  | 4.5    | 8     | café       | 5              |
| Costa Coffee    | Waterloo    | 4.1    | 8     | café       | 3              |
| The Queens Head | Denman St.  | 4.7    | 25    | bar        | 5              |

Left join

reviews.join(inspections, on=["business", "location"],           )

Left join

reviews.join(inspections, on=["business", "location"], how="left")

shape: (5, 6)
| business         | location    | review | price | type       | hygiene_rating |
| ---              | ---         | ---    | ---   | ---        | ---            |
| str              | str         | f64    | i64   | str        | i64            |
|------------------|-------------|--------|-------|------------|----------------|
| 7burgers         | Wakey Wakey | 4.2    | 15    | restaurant | 4              |
| Bang Bang Burger | Forest Rd.  | 3.8    | 12    | null       | null           |
| Costa Coffee     | City Point  | 4.5    | 8     | café       | 5              |
| Costa Coffee     | Waterloo    | 4.1    | 8     | café       | 3              |
| The Queens Head  | Denman St.  | 4.7    | 25    | bar        | 5              |

Full join

reviews.join(inspections, on=["business", "location"],                          )

Full join

reviews.join(inspections, on=["business", "location"], how="full"               )

Full join

reviews.join(inspections, on=["business", "location"], how="full", coalesce=True)

shape: (6, 6)
| business         | location    | review | price | type       | hygiene_rating |
| ---              | ---         | ---    | ---   | ---        | ---            |
| str              | str         | f64    | i64   | str        | i64            |
|------------------|-------------|--------|-------|------------|----------------|
| 7burgers         | Wakey Wakey | 4.2    | 15    | restaurant | 4              |
| Costa Coffee     | City Point  | 4.5    | 8     | café       | 5              |
| Costa Coffee     | Waterloo    | 4.1    | 8     | café       | 3              |
| The Queens Head  | Denman St.  | 4.7    | 25    | bar        | 5              |
| Wagamama         | Soho        | null   | null  | restaurant | 4              |
| Bang Bang Burger | Forest Rd.  | 3.8    | 12    | null       | null           |

Elegir una estrategia de join

2 tablas

Inner join

Representación de inner join

Left join

Representación de left join

Full join

Representación de full join

¡Vamos a practicar!

Transformación de datos con Polars