Type:	Package
Version:	0.2.0
Title:	Projection Pursuit Classification Forest
Maintainer:	Natalia da Silva <natalia.dasilva@fcea.edu.uy>
Description:	Implements projection pursuit forest algorithm for supervised classification.
License:	GPL-2 | GPL-3 [expanded from: GPL (≥ 2)]
URL:	https://github.com/natydasilva/PPforest
LazyData:	yes
Depends:	R (≥ 4.1.0)
Imports:	Rcpp (≥ 0.12.7), magrittr, plyr, dplyr (≥ 0.7.5), tidyr, doParallel, tibble, tidyselect
Suggests:	knitr, gridExtra, GGally, ggplot2, RColorBrewer, roxygen2 (≥ 3.0.0), PPtreeViz, rmarkdown
VignetteBuilder:	knitr
LinkingTo:	Rcpp,RcppArmadillo
Encoding:	UTF-8
RoxygenNote:	7.3.2
BugReports:	https://github.com/natydasilva/PPforest/issues
NeedsCompilation:	yes
Packaged:	2025-07-23 22:09:04 UTC; nataliadasilva
Author:	Natalia da Silva [aut, cre], Dianne Cook [aut], Eun-Kyung Lee [aut]
Repository:	CRAN
Date/Publication:	2025-07-23 22:20:19 UTC

NCI60 data set

Description

NCI60 data set

Usage

data(NCI60)

Format

cDNA microarrays were used to examine the variation in gene expression among the 60 cell lines. The cell lines are derived from tumors with different sites of origin. This data set contain 61 observations and 30 feature variables from 8 different tissue types.

Type

has 8 different tissue types, 9 cases of breast, 5 cases of central nervous system (CNS), 7 cases pf colon, 8 cases of leukemia, 8 cases of melanoma, 9 cases of non-small-cell lung carcinoma (NSCLC), 6 cases of ovarian and 9 cases of renal.

Gene1, Gen2, ..., Gen30

Numeric gene expression information.

A data frame with 61 rows and 31 variables

Source

Dudoit, S., Fridlyand, J. and Speed, T. P. (2002). Comparison of Discrimination Methods for the Classification of Tumors Using Gene Expression Data. Journal of the American statistical Association 97 77-87.

Predict class for the test set and calculate prediction error after finding the PPtree structure, .

Description

Predict class for the test set and calculate prediction error after finding the PPtree structure, .

Usage

PPclassify( Tree.result, test.data = NULL, Rule = 1, true.class = NULL)

Arguments

Value

predict.class predicted class

predict.error prediction error

References

Lee, YD, Cook, D., Park JW, and Lee, EK(2013) PPtree: Projection pursuit classification tree, Electronic Journal of Statistics, 7:1369-1386.

Projection Pursuit Random Forest

Description

PPforest implements a random forest using projection pursuit trees algorithm (based on PPtreeViz package).

Usage

PPforest(data, y, std = 'scale', size.tr, m, PPmethod, size.p,
 lambda = .1, parallel = FALSE, cores = 2, rule = 1)

Arguments

Value

An object of class PPforest with components.

References

da Silva, N., Cook, D., & Lee, E. K. (2021). A projection pursuit forest algorithm for supervised classification. Journal of Computational and Graphical Statistics, 30(4), 1168-1180.

Examples

#crab example with all the observations used as training
set.seed(123)
pprf.crab <- PPforest(data = crab, y = 'Type',
 std = 'no', size.tr = 0.8, m = 100, size.p = 1, 
 PPmethod = 'LDA' , parallel = TRUE, cores = 2, rule = 1)
pprf.crab

Projection pursuit classification tree with random variable selection in each split

Description

Find tree structure using projection pursuit indices of classification in each split.

Usage

PPtree_split(form, data, PPmethod='LDA', 
size.p=1,  lambda = 0.1,...)

Arguments

Value

An object of class PPtreeclass with components

References

Lee, YD, Cook, D., Park JW, and Lee, EK (2013) PPtree: Projection pursuit classification tree, Electronic Journal of Statistics, 7:1369-1386.

Examples

#crab data set

Tree.crab <- PPtree_split('Type~.', data = crab, PPmethod = 'LDA', size.p = 0.5)
Tree.crab

For each bootstrap sample grow a projection pursuit tree (PPtree object).

Description

For each bootstrap sample grow a projection pursuit tree (PPtree object).

Usage

baggtree(
  data,
  y,
  m = 500,
  PPmethod = "LDA",
  lambda = 0.1,
  size.p = 1,
  parallel = FALSE,
  cores = 2
)

Arguments

Value

data frame with trees_pp output for all the bootstraps samples.

Astralian crabs

Description

Astralian crabs

Usage

data(crab)

Format

Measurements on rock crabs of the genus Leptograpsus. The data set contains 200 observations from two species of crab (blue and orange), there are 50 specimens of each sex of each species, collected on site at Fremantle, Western Australia.

Type

is the class variable and has 4 classes with the combinations of species and sex (BlueMale, BlueFemale, OrangeMale and OrangeFemale).

FL

the size of the frontal lobe length, in mm.

RW

rear width, in mm

CL

length of midline of the carapace, in mm.

CW

maximum width of carapace, in mm.

BD

depth of the body; for females, measured after displacement of the abdomen, in mm.

A data frame with 200 rows and 6 variables

Source

Campbell, N. A. & Mahon, R. J. (1974), A Multivariate Study of Variation in Two Species of Rock Crab of genus Leptograpsus, Australian Journal of Zoology 22(3), 417 - 425.

Fish catch data set

Description

Fish catch data set

Usage

data(fishcatch)

Format

There are 159 fishes of 7 species are caught and measured. Altogether there are 7 variables. All the fishes are caught from the same lake(Laengelmavesi) near Tampere in Finland.

Type

has 7 fish classes, with 35 cases of Bream, 11 cases of Parkki, 56 cases of Perch, 17 cases of Pike, 20 cases of Roach, 14 cases of Smelt and 6 cases of Whitewish.

weight

Weight of the fish (in grams).

length1

Length from the nose to the beginning of the tail (in cm).

length2

Length from the nose to the notch of the tail (in cm).

length3

Length from the nose to the end of the tail (in cm).

height

Maximal height as % of Length3.

width

Maximal width as % of Length3.

A data frame with 159 rows and 7 variables

Source

<http://www.amstat.org/publications/jse/jse_data_archive.htm>

Glass data set

Description

Glass data set

Usage

data(glass)

Format

Contains measurements 214 observations of 6 types of glass; defined in terms of their oxide content.

Type

has 6 types of glasses.

X1

refractive index.

X2

Sodium (unit measurement: weight percent in corresponding oxide).

X3

Magnesium.

X4

Aluminum.

X5

Silicon.

X6

Potassium.

X7

Calcium.

X8

Barium.

X9

Iron.

A data frame with 214 rows and 10 variables

The image data set

Description

The image data set

Usage

data(image)

Format

Contains 2310 observations of instances from 7 outdoor images

Type

has 7 types of outdoor images, brickface, cement, foliage, grass, path, sky, and window.

X1

the column of the center pixel of the region

X2

the row of the center pixel of the region.

X3

the number of pixels in a region = 9.

X4

the results of a line extraction algorithm that counts how many lines of length 5 (any orientation) with low contrast, less than or equal to 5, go through the region.

X5

measure the contrast of horizontally adjacent pixels in the region. There are 6, the mean and standard deviation are given. This attribute is used as a vertical edge detector.

X6

X5 sd

X7

measures the contrast of vertically adjacent pixels. Used for horizontal line detection.

X8

sd X7

X9

the average over the region of (R + G + B)/3

X10

the average over the region of the R value.

X11

the average over the region of the B value.

X12

the average over the region of the G value.

X13

measure the excess red: (2R - (G + B)).

X14

measure the excess blue: (2B - (G + R)).

X15

measure the excess green: (2G - (R + B)).

X16

3-d nonlinear transformation of RGB. (Algorithm can be found in Foley and VanDam, Fundamentals of Interactive Computer Graphics).

X17

mean of X16.

X18

hue mean.

A data frame contains 2310 observations and 19 variables

Leukemia data set

Description

Leukemia data set

Usage

data(leukemia)

Format

This dataset comes from a study of gene expression in two types of acute leukemias, acute lymphoblastic leukemia (ALL) and acute myeloid leukemia (AML). Gene expression levels were measured using Affymetrix high density oligonucleotide arrays containing 6817 human genes. A data set containing 72 observations from 3 leukemia types classes.

Type

has 3 classes with 38 cases of B-cell ALL, 25 cases of AML and 9 cases of T-cell ALL.

Gene1, Gen2, ..., Gen40

gene expression levels.

A data frame with 72 rows and 41 variables

Source

Dudoit, S., Fridlyand, J. and Speed, T. P. (2002). Comparison of Discrimination Methods for the Classification of Tumors Using Gene Expression Data. Journal of the American statistical Association 97 77-87.

Lymphoma data set

Description

Lymphoma data set

Usage

data(lymphoma)

Format

Gene expression in the three most prevalent adult lymphoid malignancies: B-cell chronic lymphocytic leukemia (B-CLL), follicular lymphoma (FL), and diffuse large B-cell lym- phoma (DLBCL). Gene expression levels were measured using a specialized cDNA microarray, the Lymphochip, containing genes that are preferentially expressed in lymphoid cells or that are of known immunologic or oncologic importance. This data set contain 80 observations from 3 lymphoma types.

Type

Class variable has 3 classes with 29 cases of B-cell ALL (B-CLL), 42 cases of diffuse large B-cell lymphoma (DLBCL) and 9 cases of follicular lymphoma (FL).

Gene1, Gen2, ..., Gen50

gene expression.

A data frame with 80 rows and 51 variables

Source

Dudoit, S., Fridlyand, J. and Speed, T. P. (2002). Comparison of Discrimination Methods for the Classification of Tumors Using Gene Ex- pression Data. Journal of the American statistical Association 97 77-87.

Data structure with the projected and boundary by node and class.

Description

Data structure with the projected and boundary by node and class.

Usage

node_data(ppf, tr, Rule = 1)

Arguments

Value

Data frame with projected data for each class and node id and the boundaries

Examples

#crab data set with all the observations used as training

pprf.crab <- PPforest(data = crab, std = 'min-max', y = 'Type',
 size.tr = 1, m = 200, size.p = .5, PPmethod = 'LDA')
node_data(ppf = pprf.crab, tr = 1) 

The olive data set

Description

The olive data set

Usage

data(olive)

Format

Contains 572 observations and 10 variables

Region

Three super-classes of Italy: North, South and the island of Sardinia.

area

Nine collection areas: three from North, four from South and 2 from Sardinia.

palmitic

fatty acids percent x 100.

palmitoleic

fatty acids percent x 100.

stearic

fatty acids percent x 100.

oleic

fatty acids percent x 100.

linoleic

fatty acids percent x 100.

linolenic

fatty acids percent x 100.

arachidic

fatty acids percent x 100.

eicosenoic

fatty acids percent x 100.

A data frame contains 573 observations and 10 variables

Parkinson data set

Description

Parkinson data set

Usage

data(parkinson)

Format

A data set containing 195 observations from 2 parkinson types.

Type

Class variable has 2 classes, there are 48 cases of healthy people and 147 cases with Parkinson. The feature variables are biomedical voice measures.

X1

Average vocal fundamental frequency.

X2

Maximum vocal fundamental frequency.

X3

Minimum vocal fundamental frequency.

X4

MDVP:Jitter(%) measures of variation in fundamental frequency.

X5

MDVP:Jitter(Abs) measures of variation in fundamental frequency.

X6

MDVP:RAP measures of variation in fundamental frequency.

X7

MDVP:PPQ measures of variation in fundamental frequency.

X8

Jitter:DDP measures of variation in fundamental frequency.

X9

MDVP:Shimmer measures of variation in amplitude.

X10

MDVP:Shimmer(dB) measures of variation in amplitude.

X11

Shimmer:APQ3 measures of variation in amplitude.

X12

Shimmer:APQ5 measures of variation in amplitude.

X13

MDVP:APQ measures of variation in amplitude.

X14

Shimmer:DDA measures of variation in amplitude.

X15

NHR measures of ratio of noise to tonal components in the voice.

X16

HNR measures of ratio of noise to tonal components in the voice.

X17

RPDE nonlinear dynamical complexity measures.

X18

D2 nonlinear dynamical complexity measures.

X19

DFA - Signal fractal scaling exponent.

X20

spread1 Nonlinear measures of fundamental frequency variation.

X21

spread2 Nonlinear measures of fundamental frequency variation.

X22

PPE Nonlinear measures of fundamental frequency variation.

A data frame with 195 rows and 23 variables

Source

<https://archive.ics.uci.edu/ml/datasets/Parkinsons>

Obtain the permuted importance variable measure

Description

Obtain the permuted importance variable measure

Usage

permute_importance(ppf)

Arguments

Value

A data frame with permuted importance measures, imp is the permuted importance measure defined in Brieman paper, imp2 is the permuted importance measure defined in randomForest package, the standard deviation (sd.im and sd.imp2) for each measure is computed and the also the standardized mesure.

Examples

pprf.crab <- PPforest(data = crab, y = 'Type',
std = 'min-max', size.tr = 1, m = 100, size.p = .4, PPmethod = 'LDA', parallel = TRUE, core = 2)
permute_importance(ppf = pprf.crab) 

Global importance measure for a PPforest object as the average IMP PPtree measure over all the trees in the forest

Description

Global importance measure for a PPforest object as the average IMP PPtree measure over all the trees in the forest

Usage

ppf_avg_imp(ppf, y)

Arguments

Value

Data frame with the global importance measure

References

da Silva, N., Cook, D., & Lee, E. K. (2021). A projection pursuit forest algorithm for supervised classification. Journal of Computational and Graphical Statistics, 30(4), 1168-1180.

Examples

#crab data set with all the observations used as training

pprf.crab <- PPforest(data = crab, std = 'min-max', y = 'Type',
 size.tr = 1, m = 100, size.p = .5, PPmethod = 'LDA')
 ppf_avg_imp(pprf.crab, 'Type') 
 

Global importance measure for a PPforest object

Description

Global importance measure for a PPforest object

Usage

ppf_global_imp(data, y, ppf)

Arguments

Value

Data frame with the global importance measure

References

da Silva, N., Cook, D., & Lee, E. K. (2021). A projection pursuit forest algorithm for supervised classification. Journal of Computational and Graphical Statistics, 30(4), 1168-1180.

Examples

#crab data set with all the observations used as training
pprf.crab <- PPforest(data = crab, y = 'Type',
 std = 'no',  size.tr = 1, m = 200, size.p = .5, 
 PPmethod = 'LDA', parallel = TRUE, cores = 2)
 
ppf_global_imp(data = crab, y = 'Type', pprf.crab) 

Predict method for PPforest objects

Description

Predict method for PPforest objects

Usage

## S3 method for class 'PPforest'
predict(object, newdata, rule = 1, parallel = TRUE, cores = 2, ...)

Arguments

Value

A list with:

predtree

Matrix with individual tree predictions

predforest

Final predicted classes based on majority vote

Examples

## Not run: 
set.seed(123)
train <- sample(1:nrow(crab), nrow(crab)*.7)
crab_train <- data.frame(crab[train, ])
crab_test <- data.frame(crab[-train, ])

# if you split your data in training and test outside PPforest size.tr should be 1.
pprf.crab <- PPforest(data = crab_train, class = 'Type',
 std = 'scale', size.tr = 1, m = 200, size.p = .4, PPmethod = 'LDA', parallel = TRUE )
 
pred <- predict(pprf.crab, newdata = crab_test[,-1], parallel = TRUE) 

## End(Not run)

Print PPforest object

Description

Print PPforest object

Usage

## S3 method for class 'PPforest'
print(x, ...)

Arguments

Value

printed results for PPforest object

Data structure with the projected and boundary by node and class.

Description

Data structure with the projected and boundary by node and class.

Usage

ternary_str(ppf, id, sp, dx, dy)

Arguments

Value

Data frame needed to visualize a ternary plot

References

da da Silva, N., Cook, D. & Lee, EK. Interactive graphics for visually diagnosing forest classifiers in R. Comput Stat 40, 3105–3125 (2025). https://doi.org/10.1007/s00180-023-01323-x

Examples

#crab data set with all the observations used as training
pprf.crab <- PPforest(data = crab, std ='min-max', y = "Type",
 size.tr = 1, m = 100, size.p = .5, PPmethod = 'LDA')
 require(dplyr)
pl_ter <- function(dat, dx, dy ){
  p1  <- dat[[1]] %>% dplyr::filter(pair %in% paste(dx, dy, sep = "-") ) %>%
    dplyr::select(Class, x, y) %>%
    ggplot2::ggplot(ggplot2::aes(x, y, color = Class)) +
    ggplot2::geom_segment(data = dat[[2]], ggplot2::aes(x = x1, xend = x2,
                                               y = y1, yend = y2), color = "black" ) +
    ggplot2::geom_point(size = I(3), alpha = .5) +
    ggplot2::labs(y = " ",  x = " ") +
    ggplot2::theme(legend.position = "none", aspect.ratio = 1) +
    ggplot2::scale_colour_brewer(type = "qual", palette = "Dark2") +
    ggplot2::labs(x = paste0("T", dx, " "), y = paste0("T", dy, " ")) +
    ggplot2::theme(aspect.ratio = 1)
  p1
}
#ternary plot in tree different selected dierections
 pl_ter(ternary_str(pprf.crab, id = c(1, 2, 3), sp = 3, dx = 1, dy = 2), 1, 2 )

Obtain predicted class for new data from baggtree function or PPforest

Description

Obtain predicted class for new data from baggtree function or PPforest

Usage

trees_pred(object, xnew, parallel = FALSE, cores = 2, rule = 1)

Arguments

Value

predicted values from PPforest or baggtree

Wine data set

Description

Wine data set

Usage

data(wine)

Format

A data set containing 178 observations from 3 wine grown cultivares in Italy.

Type

Class variable has 3 classes that are 3 different wine grown cultivares in Italy.

X1 to X13

Check vbles

A data frame with 178 rows and 14 variables