In previous versions of Origin, unbalanced box plots, or rather box plots that display series of data with varying degrees of completion, would display the same indexed colors across grouped series of data regardless of whether or not all subgroups were represented within a group of data. In this tutorial we’ll show you how to use Origin 2016 to make a compelling box plot of unbalanced data with the sample data found in File: Open Sample Projects: New Graph Types in Origin 2016.
Here we see a table of different data series taken across four continents, note however that some continents have data available from one series that may be absent for another. More specifically, Europe lacks the 500 series of data while Africa has no points of data within the 100 series.
1. Let’s create a box plot of this data to see how these values compare and contrast across the series. To do this, select Plot>Statistics>Grouped Box Charts – Indexed Data… and in the prompted dialog box set your data column to column C, and your group columns to columns A and B.
2. As you can see from the preview, series 100, 300, and 500 are all represented by different colors however you’ll notice the data from Africa is offset, displaying the 300 series the same color as 100 series data. This is because the default setting is using the primary group info. (region) and color increment within the group. Double click any plot to open the Plot Details dialog. On Group tab and take a look of the Subgrouping settings.
3. To alter the color scheme of our box plot, while still in the Group tab, change the Increment column value next on the Box Color line to By One, then right click on the color palette displayed next to the Box Color line and select Load Increment List to change the color scheme used on the graph. Then under the Pattern tab, change the border color to black, and click okay.
4. Your graph should appear as it does below. Our data is displayed correctly, but let’s clean up the formatting a little more to make this a publication quality graph.
To change the tick symbols to match those of the sample graph, open the Plot Details dialog and under the percentile tab uncheck the Max, 99%, 1%, and Min tickboxes. For the mean let’s set our symbol to a hollow circle, increase its size to 8 and change the Fill Color to None.
5. In addition to our x-axis labels, vertical axis grids can also help set apart our different data. To activate these, double click on any the x-axis to prompt the X-Axis dialog box. In the Grids tab check off the Major and Minor Grid Lines tickboxes and, in this case, we’ll also change the minor grid line style to solid.
5. This graph still has a somewhat busy x-axis. If you want to remove the cell borders for the data series axis (100, 300, 500), open the Tick Labels tab of the X-Axis dialog. The settings for the axis ticks of the data series will be under the Table tab of the Bottom 2 menu. Here we uncheck Auto, as well as the Inside and Outside Border tickboxes, removing the borders around our data series labels. To add a gray fill to our continents label, select Bottom 1 and in the Table tab change the fill color to LT Gray.
6. To replicate our graph to a tee, our y-axis values will need to be modified from their default range. To do this double click on the y-axis to open the Y-Axis dialog, and change the From value to -3.5 and the To value to 2.5.
Italicize the series numbers, add a title, and you should have a box plot that appears as the one below.