Geometric Layers: Choosing the Visual Representation

📊 The Final Ingredient

In the previous lessons, we covered the plot’s backbone — your data, and the aesthetics that map it to visual properties. But if you tried running a plot with just those, you’ll have noticed: nothing shows up inside the panel (a.k.a. the gray area).

That’s because every ggplot needs at least one layer to actually display anything. And most often, that layer is a geom_*().

So far, you’ve already seen and used several of these layers:

🧩 Not All Geoms Want the Same Things

All of the geoms listed above will happily show up when added to our "empty panel" code with just x and y mapped — well, all but one.

 Can you guess which one won’t play along? 

Yes, this task should ring a bell if you explored the “Lost for Words” exercise in the Aesthetics: Caveats lesson.

It’s geom_text() that stays silent. Not because of a bug — but because it’s still waiting for one more piece: a label aesthetic. While we can draw a point or bar with two coordinates, geom_text() and geom_label() need to know what to write.

This isn’t unique to geom_text(). For example, geom_point() can’t draw a point without knowing where to place it, so it expects both x and y. Otherwise ggplot2 will throw an error and inform you about missing aesthetics:

Error in `geom_point()`: Problem while setting up geom. `geom_point()` requires the following missing aesthetics: y.

 Every layer expects certain aesthetics to be present. 

• Geometries that require two positions, usually x and y, are the most common case.
  
  
• Others need a third aesthetic such as label for geom_text() or fill for geom_tile().
  
  
• Some layers are minimalists: for example, geom_bar() only required an x or y position to place the bar — and calculates the height or width, respectively, itself.
  
  
• And some demand many more, often specific aesthetics such as additional xend and yend to draw a line with geom_segment() to define start and end coordinates.
  
  

Now that you know some geoms need different numbers and types of aesthetics, let’s explore them in more detail.

📐 A World of Chart Types Geoms

The world of data visualization includes countless chart types. While the possibilities are virtually endless, mastering the fundamental ones is key to clear and effective communication.

Click any chart icon to see which geom it uses and what aesthetics are required — or explore the other way around!

DistributionCorrelationRankingPart Of A wholeEvolutionMapFlow

Violin

Density

Histogram

Boxplot

Ridgeline

🥞 Layer by Layer

One of the most powerful ideas in ggplot2 is that plots are built layer by layer. That means you’re not limited to a single geom — you can stack multiple geoms on top of the same data, or even different data.

For example, you might want to: