The FFTrees package creates fast and frugal decision trees (FFTs). Fast and frugal decision trees are very simple heuristics decision algorithms that, in contrast to regression and more complex algorithms such as random forests and (non-frugal) decision trees, allow people to make fast decisions based on minimal information. FFTs are not just easy to use, they can also predict data as well, if not better, than more complex algorithms such as regression (Gigerenzer and Todd 1999; Martignon, Katsikopoulos, and Woike 2008). Because they are so effective and easy to use and, fast and frugal trees have been used to improve decision making in medical (Green and Mehr 1997; Jenny et al. 2013) and financial (Neth et al. 2014) contexts.

The purpose of FFTrees is to allow anyone to quickly and easily generate, visualise, and use FFTs for their own data.

# Install FFTrees from CRAN (only necessary once)
install.packages("FFTrees")

library(FFTrees)

heart.fft <- FFTrees(formula = diagnosis ~.,        # Criterion
                     data = heart.train,            # Training data   
                     data.test = heart.test)        # Testing data

plot(heart.fft,                                     # FFTrees object
     main = "Heart Disease",                        # Plot title
     decision.names = c("Healthy", "Diseased"),     # Decision names
     data = "test")                                 # Show statistics for test data