09-06-layers-position-adjustments.qmd

# Layers

# Position Adjustments

```{r}
#| output: false

library(tidyverse)
```


> 1. What is the problem with the following plot? How could you improve it?

```{r}
ggplot(mpg, aes(x = cty, y = hwy)) +
  geom_point()
```

I'm gonna say sight unseen that we have a really small variance in x, meaning
that we have overlapping values.

Boy was I right. Let's add some transparency:

```{r}
ggplot(mpg, aes(x = cty, y = hwy)) +
  geom_point(alpha = 0.6, color = "#6994d1")
```

Now let's jitter the points:

```{r}
ggplot(mpg, aes(x = cty, y = hwy)) +
  geom_point(alpha = 0.6, color = "#6994d1", position = "jitter")
```

> 2. What, if anything, is the difference between the two plots? Why?

```{r}
ggplot(mpg, aes(x = displ, y = hwy)) +
  geom_point()

ggplot(mpg, aes(x = displ, y = hwy)) +
  geom_point(position = "identity")
```

Oh I bet these are the same. I'm almost certain the default value for
`position` with a scatter plot is "identity".

> 3. What parameters to `geom_jitter()` control the amount of jittering?

`width` and `height`. They are proportions of the resolution of the data, which
is to say the differences between them. I think that means that for more
continuous data, they move less.

> 4. Compare and contrast `geom_jitter()` with `geom_count()`.

Well, one shows overplotting on a scatter by moving, and the other by
increasing the size of the points.

> 5. What’s the default position adjustment for `geom_boxplot()`? Create a
>    visualization of the mpg dataset that demonstrates it.

The default value is `position_dodge2`.

```{r}
ggplot(mpg) +
  geom_boxplot(
    aes(x = cty, y = hwy, group = cyl),
    position = "identity"
  )
```