Making a 3D Scatter Plot
Most graphs have two axes, referred to often as the $ x $ axis and the $ y $ axis. We use a graph to understand the relationship between these two variables. A three-dimensional graph lets you introduce a third axis, typically called the $ z $ axis, and can help you understand the relationship between three variables.
Three-dimensional scatter plots are best when you have at least three features that you want to highlight in your graph. Using color and size you can highlight two additional features in your graph. Plotly’s interactive tools also let you explore your plot by zooming, hovering, and rotating your plot.
For this tutorial, we’re using the “Prestige” dataset from R’s “car” package. The data set contains information about status (quantified with the Pineo-Porter index), salary, years of schooling, and gender for job categories in the 1970s. You can learn more in this document. If you follow along with the R tutorial at the link, you can take advantage of Plotly’s awesome styling, collaboration and sharing tools. If you want to know about integrating Plotly with R, we have excellent documentation!
Step 1: Set up the grid
|For this tutorial, go to: https://plot.ly/~mariahh/85. Click on Fork and edit to open this data set in your workspace.|
|Select 3d scatter plots from the MAKE A PLOT menu. We’re going to compare education, income and prestige. We’ll put “income (1971)” on the x axis by selecting choose as x and “education” (this is the number of years of education required for a certain job) on the y axis by selecting choose as y. Finally, we put “Pineo-Porter prestige score” on the z axis by selecting choose as z.|
|We can assign color as a variable for “gender” in our 3d plot with the Group by option, located in the sidebar. Select choose as G in the column labeled “gender”. Click the blue 3d scatter plot button, in the sidebar, to create the chart. This will automatically save it and prompt you with the option to share it. If you make your graph public, you can share it via Facebook, Twitter or Google+. Need help embedding your graph in a website or blog? We have a tutorial for that.|
Step 2: Style it
|Your plot should look something like this.|
|Open the TRACES popover in the toolbar. Using the Style tab you can change the style, thickness, and color of the data points. Here we’ve made the jobs dominated by women blue and those dominated by men green. To close the Traces popover, click the X in the upper right.|
|You can add a title by clicking on Click to enter Plot title. You can change the font and size of the title and other annotations using the LAYOUT popover, and Global Font.|
|Still in the LAYOUT popover, we can change the Scene color, the margin color, and more.|
Step 3: It’s interactive!
|Your 3d scatter plot is interactive! Different views help us understand different information contained in the data set. In this view, we can see information about prestige and years of schooling. We can see that the number of years of schooling is correlated to prestige. This relationship appears to hold true for jobs dominated by both men (in green) and women (in blue).|
|Changing the view allows us to compare salary and schooling. In this view we see clearly that jobs dominated by women tend to be lower wage jobs.|