 Raw also provides several of the more obscure variations on scatter plots. One of these is the convex hull, where convex means a shape that never goes in. It's never a concave that's drawn to surround the points. I'm going to come down here, and this time we only get to choose two data dimensions. I'm using the movie data set again, so I'll just bring in rating on IMDB on the X, and total domestic box office, the earnings on the Y. And each of these points represents one of the movies. Now, what's interesting is we can't label it, so it won't tell us what they are. Hopefully we've made another diagram, which will tell us which ones they are. But what this does is it draws a shape that is flat or convex. It's never concave, it doesn't bend in. Which gives us an idea of the general shape of the collection of points. And so you can see that it is still basically uphill, and it's kind of long, kind of narrow. And this can be a way of getting the overall shape as well as a sense of groupings or possibly outliers within the data.