Graphing is a vital part of any data analysis project. Graphs visually reveal patterns and relationships between variables and provide invaluable information. At times, the patterns may be interesting; however, the scaling of the data can simultaneously interfere with the messages to be conveyed.
When units and scale vary greatly, seeing detail in all variables on a plot becomes quite impossible. This is when you know your multi-variable plot needs multiple, varying scales. Let’s look at our options…
Double Y Plots
Many graphing tools have a Graph type option called Double-Y. This graph type makes it possible for you to select one or more variables associated with the left Y axis and one or more variables to associate with the right Y axis. This is a simple way of creating a compound graph that shows variables with two different scales.
For example, open the STATISTICA data file, Baseball.sta, from the path C:/Program Files/StatSoft/STATISTICA 12/Examples/Datasets. Several of the variables in this example data file have very different scales.
On the Graphs tab in the Common group, click Scatterplot. In the 2D Scatterplots Start Panel, select the Advanced tab. In the Graph type group box, select Double-Y.
Now, click the Variables button, and in the variable selection dialog box, select RUNS as X, WIN as Y Left, and DP as Y Right. Click the OK button.
Click OK in the 2D Scatterplots Startup Panel to create the plot. The result lists the two Y variables with separately determined scales.
WIN shows a scale from 0.25 to 0.65. This is the season winning proportion. The variable DP is shown on a scale from 100 to 220 and is the number of double plays in the season. Because of the great difference in the scale of these two variables, a Double-Y plot is the best way to simultaneously show these variables’ relationships with the X factor, RUNS.
Multiple Y Plots
An additional option is available for creating plots with multiple axis scales. This option is used when you need more scales than the Double-Y allows or when you need an additional axis in another place or capacity.
Continuing the same example, add a second variable, BA, to the Y Left variable list.
Click OK to create the new plot.
Now, WIN and BA share the left Y axis. BA, batting average, is on a scale of .2 to .3. Giving BA a separate Y axis scale would show more detail in the added variable. To do this, right-click in the graph, and on the shortcut menu select Graph Options. Select the Axis – General tab of the Graph Options dialog box.
From the Axis drop-down menu, select the Y left axis. Then click Add new axis. A new Y left axis is added to the plot called Y left’.
Next, the BA variable needs to be related to that axis and customized. Select the Plot – General tab to make this change.
On the Plot drop-down list, select the variable BA. Then, in the Assignment of axis group, select the Custom option button, and specify Y left’ as the custom axis.
Click OK to update the plot.
The resulting plot now has three Y variables plotted, each with its own Y axis scaling and labeling. Showing patterns and relationships in data of varying scale is made easy with multiple axes.