FFTrees 1.6.6

The goal of FFTrees is to create and visualize fast-and-frugal decision trees (FFTs) from data with a binary outcome following the methods described in Phillips, Neth, Woike & Gaissmaier (2017).

Installation

You can install the released version of FFTrees from CRAN with:

install.packages("FFTrees")

And the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("ndphillips/FFTrees", build_vignettes = TRUE)

Examples

library(FFTrees)
#> 
#>    O
#>   / \
#>  F   O
#>     / \
#>    F   Trees 1.6.6
#> 
#> Email: Nathaniel.D.Phillips.is@gmail.com
#> FFTrees.guide() opens the main guide.

Let’s create a fast-and-frugal tree predicting heart disease status (“Healthy” vs. “Diseased”) based on a heart.train dataset, and test it on heart.test a testing dataset.

Here are the first new rows and columns of our datasets. The key column is diagnosis, a logical column (TRUE and FALSE) which indicate, for each patient, whether or not they have heart disease.

Here is heart.train (the training dataset) which contains data from 150 patients:

heart.train
#> # A tibble: 150 × 14
#>    diagnosis   age   sex cp    trestbps  chol   fbs restecg     thalach exang
#>    <lgl>     <dbl> <dbl> <chr>    <dbl> <dbl> <dbl> <chr>         <dbl> <dbl>
#>  1 FALSE        44     0 np         108   141     0 normal          175     0
#>  2 FALSE        51     0 np         140   308     0 hypertrophy     142     0
#>  3 FALSE        52     1 np         138   223     0 normal          169     0
#>  4 TRUE         48     1 aa         110   229     0 normal          168     0
#>  5 FALSE        59     1 aa         140   221     0 normal          164     1
#>  6 FALSE        58     1 np         105   240     0 hypertrophy     154     1
#>  7 FALSE        41     0 aa         126   306     0 normal          163     0
#>  8 TRUE         39     1 a          118   219     0 normal          140     0
#>  9 TRUE         77     1 a          125   304     0 hypertrophy     162     1
#> 10 FALSE        41     0 aa         105   198     0 normal          168     0
#> # … with 140 more rows, and 4 more variables: oldpeak <dbl>, slope <chr>,
#> #   ca <dbl>, thal <chr>

Here is heart.test (the testing / prediction dataset) which contains data from a new set of 153 patients:

heart.test
#> # A tibble: 153 × 14
#>    diagnosis   age   sex cp    trestbps  chol   fbs restecg     thalach exang
#>    <lgl>     <dbl> <dbl> <chr>    <dbl> <dbl> <dbl> <chr>         <dbl> <dbl>
#>  1 FALSE        51     0 np         120   295     0 hypertrophy     157     0
#>  2 TRUE         45     1 ta         110   264     0 normal          132     0
#>  3 TRUE         53     1 a          123   282     0 normal           95     1
#>  4 TRUE         45     1 a          142   309     0 hypertrophy     147     1
#>  5 FALSE        66     1 a          120   302     0 hypertrophy     151     0
#>  6 TRUE         48     1 a          130   256     1 hypertrophy     150     1
#>  7 TRUE         55     1 a          140   217     0 normal          111     1
#>  8 FALSE        56     1 aa         130   221     0 hypertrophy     163     0
#>  9 TRUE         42     1 a          136   315     0 normal          125     1
#> 10 FALSE        45     1 a          115   260     0 hypertrophy     185     0
#> # … with 143 more rows, and 4 more variables: oldpeak <dbl>, slope <chr>,
#> #   ca <dbl>, thal <chr>

Now let’s use FFTrees() to create a fast and frugal tree from the heart.train data and test their performance on heart.test

# Create an FFTrees object from the heartdisease data
heart.fft <- FFTrees(formula = diagnosis ~., 
                     data = heart.train,
                     data.test = heart.test, 
                     decision.labels = c("Healthy", "Disease"))
#> Setting goal = 'wacc'
#> Setting goal.chase = 'waccc'
#> Setting cost.outcomes = list(hi = 0, mi = 1, fa = 1, cr = 0)
#> Growing FFTs with ifan
#> Fitting other algorithms for comparison (disable with do.comp = FALSE) ...

# See the print method which shows aggregatge statistics
heart.fft
#> FFTrees 
#> - Trees: 7 fast-and-frugal trees predicting diagnosis
#> - Outcome costs: [hi = 0, mi = 1, fa = 1, cr = 0]
#> 
#> FFT #1: Definition
#> [1] If thal = {rd,fd}, decide Disease.
#> [2] If cp != {a}, decide Healthy.
#> [3] If ca <= 0, decide Healthy, otherwise, decide Disease.
#> 
#> FFT #1: Prediction Accuracy
#> Prediction Data: N = 153, Pos (+) = 73 (48%) 
#> 
#> |         | True + | True - |
#> |---------|--------|--------|
#> |Decide + | hi 64  | fa 19  | 83
#> |Decide - | mi 9   | cr 61  | 70
#> |---------|--------|--------|
#>             73       80       N = 153
#> 
#> acc  = 81.7%  ppv  = 77.1%  npv  = 87.1%
#> bacc = 82.0%  sens = 87.7%  spec = 76.2%
#> E(cost) = 0.183
#> 
#> FFT #1: Prediction Speed and Frugality
#> mcu = 1.73, pci = 0.87

# Plot the best tree applied to the test data
plot(heart.fft,
     data = "test",
     main = "Heart Disease")

# Compare results across algorithms in test data
heart.fft$competition$test
#>   algorithm   n hi fa mi cr      sens   spec    far       ppv       npv
#> 1   fftrees 153 64 19  9 61 0.8767123 0.7625 0.2375 0.7710843 0.8714286
#> 2        lr 153 55 13 18 67 0.7534247 0.8375 0.1625 0.8088235 0.7882353
#> 3      cart 153 50 19 23 61 0.6849315 0.7625 0.2375 0.7246377 0.7261905
#> 4        rf 153 58  8 15 72 0.7945205 0.9000 0.1000 0.8787879 0.8275862
#> 5       svm 153 55  7 18 73 0.7534247 0.9125 0.0875 0.8870968 0.8021978
#>         acc      bacc      cost cost_decisions cost_cues
#> 1 0.8169935 0.8196062 0.1830065      0.1830065         0
#> 2 0.7973856 0.7954623 0.2026144      0.2026144        NA
#> 3 0.7254902 0.7237158 0.2745098      0.2745098        NA
#> 4 0.8496732 0.8472603 0.1503268      0.1503268        NA
#> 5 0.8366013 0.8329623 0.1633987      0.1633987        NA

Because fast-and-frugal trees are so simple, you can create one ‘from words’ and apply it to data!

For example, below we’ll create a tree with the following 4 nodes and evaluate its performance on the heart.test data

1. If sex = 1, predict Disease.
2. If age < 45, predict Healthy.
3. If thal = {fd, normal}, predict Disease.
4. Otherwise, predict Healthy

# Create your own custom FFT 'in words' and apply it to data

# Create my own fft
my.fft <- FFTrees(formula = diagnosis ~., 
                  data = heart.train,
                  data.test = heart.test, 
                  decision.labels = c("Healthy", "Disease"),
                  my.tree = "If sex = 1, predict Disease.
                             If age < 45, predict Healthy.
                             If thal = {fd, normal}, predict Disease. 
                             Otherwise, predict Healthy")
#> Setting goal = 'wacc'
#> Setting goal.chase = 'waccc'
#> Setting cost.outcomes = list(hi = 0, mi = 1, fa = 1, cr = 0)
#> Fitting other algorithms for comparison (disable with do.comp = FALSE) ...

# Plot my custom fft and see how it did
plot(my.fft,
     data = "test",
     main = "Custom FFT")

As we can see, the tree has very high sensitivity but terrible specificity.

Citation

APA Citation

Phillips, Nathaniel D., Neth, Hansjoerg, Woike, Jan K., & Gaissmaier, W. (2017). FFTrees: A toolbox to create, visualize, and evaluate fast-and-frugal decision trees. Judgment and Decision Making, 12(4), 344-368.

We had a lot of fun creating FFTrees and hope you like it too! We have an article introducing the FFTrees package in the journal Judgment and Decision Making titled FFTrees: A toolbox to create, visualize,and evaluate fast-and-frugal decision trees. We encourage you to read the article to learn more about the history of FFTs and how the FFTrees package creates them.

If you use FFTrees in your work, please cite us and spread the word so we can continue developing the package

Here are some example publications that have used FFTrees (find the full list at Google Scholar)

• Lötsch, J., Haehner, A., & Hummel, T. (2020). Machine-learning-derived rules set excludes risk of Parkinson’s disease in patients with olfactory or gustatory symptoms with high accuracy. Journal of Neurology, 267(2), 469-478.
• Kagan, R., Parlee, L., Beckett, B., Hayden, J. B., Gundle, K. R., & Doung, Y. C. (2020). Radiographic parameter-driven decision tree reliably predicts aseptic mechanical failure of compressive osseointegration fixation. Acta Orthopaedica, 91(2), 171-176.
• Klement, R. J., Sonke, J. J., Allgäuer, M., Andratschke, N., Appold, S., Belderbos, J., … & Mantel, F. (2020). Correlating dose variables with local tumor control in stereotactic body radiotherapy for early stage non-small cell lung cancer: A modelling study on 1500 individual treatments. International Journal of Radiation Oncology* Biology* Physics.
• Nobre, G. G., Hunink, J. E., Baruth, B., Aerts, J. C., & Ward, P. J. (2019). Translating large-scale climate variability into crop production forecast in Europe. Scientific reports, 9(1), 1-13.
• Buchinsky, F. J., Valentino, W. L., Ruszkay, N., Powell, E., Derkay, C. S., Seedat, R. Y., … & Mortelliti, A. J. (2019). Age at diagnosis, but not HPV type, is strongly associated with clinical course in recurrent respiratory papillomatosis. PloS one, 14(6).