Scatter Plot

Making a Scatter Plot

The next chart we will be learning to create would be a scatter plot. Scatter plots are used to observe relationships between two variables. An example of this would be comparing the heights and diameters of trees, where the position of each dot correlates to that specified height and diameter. The data as a whole can represent a relationship: strong positive/negative linear, moderate positive/negative linear, or no relationship. As a data scientist, one of the most important patterns is seeing how points cluster from other points, if there are any gaps within the dataset, and identifying outliers for within the set. We focus on these aspects to make predictions on future datasets and understand trends.

This next dataset we are using to create a scatter plot is about dog breeds and their traits. The chart will compare maximum life spans to maximum weights to see how they're related.

To follow along, we can download the dogs dataset here.

Here is a snippet of what the dataset should look like:

Loading and Formatting

As mentioned previously, to load and read in the dataset, we will need to create a DataFrame component named frame. Using the frame, we must use the Load function and type in the file path of the dogs CSV. Recall that a CSV is a comma separated text file that holds in data.

use Libraries.Compute.Statistics.DataFrame
use Libraries.Interface.Controls.Charts.ScatterPlot
            
DataFrame frame
frame:Load("data/Dogs.csv")

Note that, we stored this dataset in a data folder.

Once the data has been loaded in, we will now extract this data to use in the data chart. We will be using two functions from our frame component, AddSelectedColumns(text heading) and AddSelectedFactors(text heading). In this instance, because scatter plots need two data variables for comparisons, we will be extracting the maximum life span and weight for the dogs using our AddSelectedColumns(text header) function. We will also be extracting dog breeds with our AddSelectedFactors(text header) function to discriminate the two variables through the dots. The usage of these two functions are shown below:

We should have the following code:

// pull out selected data
frame:AddSelectedFactors("Breed Group")
frame:AddSelectedColumns("Maximum Life Span")
frame:AddSelectedColumns("Maximum Weight")

Now it is time to create the Scatter Plot which can be done with the following code. This creates a chart object from our DataFrame component, frame. The rest of this lesson, we will be using the chart object we have created to change and format the rest of our scatter plot.

// using the data frame, format data by creating a scatter plot chart component
ScatterPlot chart = frame:ScatterPlot()
chart:Display()

Calling the Display() function will give us a pop-up of our formatted data so far. We still need to give meaning to our data, therefore, the following steps will show us how to label and customize our chart.

Labeling the Scatter Plot

In order for viewers to understand our data, labels give a clear comprehension of what is being presented. This means that we will be labeling the x axis, y axis, legend, and giving our chart a title that describes the dataset. To do so, we will call the following functions with our "chart" object: SetTitle(text name), SetXAxisTitle(text name), SetYAxisTitle(text name), SetLegendTitle(text name), and SetSubtitle(text name). Here is a brief description on what each function does and what it takes in.

// label your scatter plot
chart:SetXAxisTitle"Maximum Life Span (years)")
chart:SetYAxisTitle("Maximum Weight (pounds)")
chart:SetLegendTitle("Breed Group")
chart:SetSubtitle("Does weight correlate to life span for dogs?")
chart:SetTitle("Dog Weight and Life Span")

Note, if we would like to see the data chart so far, we can type "chart:Display()" to view it with the labels we created.

Customizing the Scatter Plot

Now that we have our data labeled, we can customize our data to our liking, such as adjusting the intervals, changing starting values, and changing the color. We will be playing around with all these features and to do so, we will be again, using our chart object to call these functions. The functions we will be using for this would be: SetLegendLocationToBottom(), SetColorPaletteToDisurbing(), SetFontSize(integer size), FlipOrientation(), and ShowLinearRegression(bool). Here are brief descriptions on what each function does and how to use it.

// set the legend location, choices are left, right, top and bottom
chart:SetLegendLocationToBottom()

// color palette contains yellows, oranges, browns, and greens
chart:SetColorPaletteToDisturbing()

// adjust font size by preference, here we set it to 30 pt
chart:SetFontSize(30)

// if we would like to switch the x and y axis
chart:FlipOrientation()

Displaying The Chart

Congratulations, our Scatter Plot is constructed! Now we can display the chart with the Display() function. There are two ways to do this, letting it automatically display and specifying a specific window size. By doing chart:Display() it will display in a size equal to the screen size. By doing chart:Display(num, num), it will display the chart in a respected constraint window size. We will be using the specified display.

chart:Display(1000, 750)

Now, feel free to clean, build, and run our program and we shortly should see a Game window pop-up. This is our Scatter Plot! To view the entire code, click here to view the file.

Try it Yourself!

Press the blue run button to execute the code in the code editor. Press the red stop button to end the program. Your program will work when the console outputs "Build Successful!"

Final Chart

Further Useful Scatter Plot Functions

To view more examples with charts, we can reference the Quorum Curriculum Repository for charts.

Next Tutorial

In the next tutorial, we will discuss box plot, which describes how to use the box plot chart in quorum.