XLSTAT: 3D plot in Excel

 

Background

Using a 3DPlot helps to visualize data in 3 dimensions.

 

Dataset for Example

The (attached) data correspond to the outputs (rows points and columns points) of correspondence analysis.

 

Setting Up

• Open XLSTAT.

• Select the data you want to use for the 3D visualization.

  Note that the sum of the contributions on the 3 axes is included. It can be used to determine the size of the bubbles on the 3D map.

• Click on the Visualizing data menu and select the XLSTAT-3DPlot feature:

• Select Table and click  OK .

 

• The XLSTAT-3DPlot automatically draws a plot out of the data you selected.

Notice that XLSTAT chose a scatter plot, but many other types of plots are available (check the  Charts tab). Starting from this plot, you can manipulate several display options to reach the result that fits your needs.

We will manipulate axes, objects (points), annotations as well as lattices.

 

Axes Configuration

• Click on the Axes tab.

• Make sure you select the F1, F2 and F3 data under the X, Y and Z axes, respectively. Note: By default, XLSTAT-3DPlot chooses the three first columns in the dataset (Labels, F1 and F2 in our case) to build the X, Y and Z axes. We would like to switch to F1, F2 and F3 instead.

• Uncheck the Data labels option as they are not very important in the context of correspondence analysis.

 
Resulting Plot:

• You may modify other options related to the axes (arrow type, labels etc.…).

 

Object Manipulation

Coloration : It would be interesting to color points according to the data column “type”.

• Go to the Object tab.

• Under the Color option (top left), select the Type column.

Size : You may want to let the point size be proportional to the Sum of the contributions column data.

• Select this column in the corresponding slot in the objects tab.

• Notice that doing so makes the interpretation in perspective a bit more complicated. Instead you can use filtering to remove data points associated to a sum of the contributions lower than a certain threshold.

Resulting Plot:

Annotations

For this example we will use the names stored in the Labels columns to tag the points.

• Go to the Annotations  tab.

• Select the Labels column under the Labels menu.

• Increase the font size for a better visibility.

Resulting Plot:

 

Lattices

Lattices allow a more comfortable navigation in 3D. We will add a lattice at the base (axes F1/F2) and remove cumbersome annotations.

 
Resulting plot:

 

Final Steps (filtering...)

Finally, many more options are available.

For example you may use filtering to only keep data points that have a sum of contributions higher than, say, 40%:

 
The mouse wheel allows you to move the image back and forth.

Rotations can be done by clicking on the right button and by moving the mouse to the left and to the right (while keeping the button down).

Once you are satisfied with the result, you can copy the image (use the Ctrl C shortcut) and paste it in another application (Word or Powerpoint for example), or to save the visualization for future use or modification.