Box Plots

Box plots are useful for comparing the probability distributions of two or more simulations. They are especially useful for comparing yield distributions for each phase of the SOI.

Box plots rank the data in a simulation from the highest to the lowest value, and show selected percentiles of each distribution. Percentiles indicate the proportion of years in a simulation predicted below a certain value. For example, in 9 years out of 10 a lower value was predicted than the 90th percentile, compared to 25 years out of 100 for the 25th percentile.

To analyse simulations using box plots:

1. Open the [Charts] menu and select [Box plots…],
   …or click on the icon on the toolbar.
2. Select the data for analysis in the Box plots form. Choose the simulation(s) you wish to analyse in the upper selection window, and the data you wish to analyse in the lower window.

The Percentiles window in the lower part of the screen allows you to alter the percentiles represented by the different parts of the box plot. An alternative is to replace the whiskers of the box and whisker plot with extended boxes (shown in green).