coordinates.Rmd

---
title: Coordinates
---

```{r}
library(gglite)
```

Coordinate systems control how the positional channels (x and y) are
interpreted. Use `coordinate()` to set the type and options.

## Cartesian (default)

The default coordinate system. Typically you don't need to specify it
explicitly, but you can do so for clarity.

```{r}
df = data.frame(x = c('A', 'B', 'C', 'D'), y = c(3, 7, 2, 5))
g2(df, x = 'x', y = 'y') |>
  mark_interval() |>
  coordinate('cartesian')
```

## Polar

Maps x to angle and y to radius. Useful for rose charts and radar-like
displays.

```{r}
g2(df, x = 'x', y = 'y') |>
  mark_interval() |>
  coordinate('polar')
```

### Polar with innerRadius (donut rose)

```{r}
g2(df, x = 'x', y = 'y', color = 'x') |>
  mark_interval() |>
  coordinate('polar', innerRadius = 0.4)
```

### Polar with custom start/end angles

```{r}
g2(df, x = 'x', y = 'y', color = 'x') |>
  mark_interval() |>
  coordinate('polar', startAngle = -pi / 2, endAngle = pi / 2)
```

## Theta

Maps data values to angular extent. Used for pie and donut charts.

```{r}
g2(df, x = 'x', y = 'y', color = 'x') |>
  mark_interval() |>
  transform_of('stackY') |>
  coordinate('theta')
```

### Donut chart (theta with innerRadius)

```{r}
g2(df, x = 'x', y = 'y', color = 'x') |>
  mark_interval() |>
  transform_of('stackY') |>
  coordinate('theta', innerRadius = 0.5)
```

## Radial

Maps y to the radial direction. Suitable for radial bar charts.

```{r}
g2(df, x = 'x', y = 'y', color = 'x') |>
  mark_interval() |>
  coordinate('radial')
```

### Radial with innerRadius

```{r}
g2(df, x = 'x', y = 'y', color = 'x') |>
  mark_interval() |>
  coordinate('radial', innerRadius = 0.3)
```

## Parallel

Maps multiple numeric variables to parallel axes. Use a `position`
encoding (a character vector of column names) instead of `x`/`y`.

```{r}
g2(iris, position = c('Sepal.Length', 'Sepal.Width',
    'Petal.Length', 'Petal.Width'), color = 'Species') |>
  mark_line() |>
  coordinate('parallel') |>
  legend_of('color', position = 'bottom') |>
  padding_of(top = 30)
```

## Radar

Displays data on radial axes emanating from a center point. A radar chart
is a line or area chart in polar coordinates. Use long-format data with
`x` (category), `y` (value), and `color` (series) encodings. All values
should be on the same scale (e.g., 0--100).

```{r}
df_radar = data.frame(
  item = rep(c('Design', 'Dev', 'Marketing', 'Sales', 'Support'), 2),
  score = c(80, 90, 65, 75, 85, 60, 70, 85, 80, 70),
  team = rep(c('A', 'B'), each = 5)
)
g2(df_radar, x = 'item', y = 'score', color = 'team') |>
  mark_area(style = list(fillOpacity = 0.5)) |>
  mark_line(style = list(lineWidth = 2)) |>
  mark_point() |>
  coordinate('polar') |>
  scale_of('x', padding = 0.5, align = 0) |>
  scale_of('y', domainMin = 0, domainMax = 100) |>
  axis_of('y', grid = TRUE, title = FALSE)
```

## Helix

Arranges data along a helix spiral. Works best with `mark_interval()`.

```{r}
df_helix = data.frame(x = paste0('D', 1:50), y = abs(sin(1:50 / 5)))
g2(df_helix, x = 'x', y = 'y', color = 'y') |>
  mark_interval() |>
  coordinate('helix')
```

## Transpose (coord_flip)

Swap x and y axes, equivalent to ggplot2's `coord_flip()`. This is a
coordinate transform, not a type.

```{r}
df = data.frame(x = c('A', 'B', 'C', 'D'), y = c(3, 7, 2, 5))
g2(df, x = 'x', y = 'y') |>
  mark_interval() |>
  coord_transpose()
```

### Transpose with parallel

```{r}
g2(iris, position = names(iris)[-5], color = 'Species') |>
  mark_line() |>
  coordinate('parallel') |> 
  coord_transpose()
```