dataframes-columns.qmd

---
# jupyter: julia-1.10
engine: julia
---

# Operations on columns

```{julia}
using DataFrames, PalmerPenguins
using Tidier
import DataFramesMeta as DFM

penguins = PalmerPenguins.load() |> DataFrame;
@slice_head(penguins, n = 10)
```

## Selecting (or: throwing columns away)

### Selecting `n` columns

**Problem:** Select only some columns.

::: {.panel-tabset}

## Tidier

```{julia}
@select penguins species body_mass_g
```

## DataFramesMeta

```{julia}
DFM.@select penguins :species :body_mass_g
```

## DataFrames

```{julia}
DFM.select(penguins, [:species, :body_mass_g])
```

:::

### Selecting columns from a variable

**Problem:** Select only some columns whose names are stored in a variable.

::: {.panel-tabset}

```{julia}
my_columns = [:species, :body_mass_g];
```

## Tidier

```{julia}
@eval @select penguins $my_columns...
```

## DataFramesMeta

```{julia}
DFM.@select penguins $my_columns
```

## DataFrames

```{julia}
DFM.select(penguins, my_columns)
```

:::

## Mutating (or: creating columns)

### Creating one column based on another one

**Problem:** Create the column `body_mass_kg` by dividing `body_mass_g` by 1000.

::: {.panel-tabset}

## Tidier

```{julia}
@mutate penguins body_mass_kg = body_mass_g / 1000
```

## DataFramesMeta

```{julia}
DFM.@rtransform penguins :body_mass_kg = :body_mass_g / 1000
```

## DataFrames

```{julia}
penguins2 = copy(penguins);
penguins.body_mass_kg = penguins.body_mass_g ./ 1000;
penguins2
```

:::

## Conditionally mutating columns