Nothing
R/regressors.R
In less: Learning with Subset Stacking
####################
# HELPER CLASSES
####################
#' @title BaseEstimator
#'
#' @description A dummy base R6 class that provides get_all_fields, get_attributes and set_random_state functionalities for estimators
#'
#' @return R6 Class of BaseEstimator
#' @export
BaseEstimator <- R6::R6Class(classname = "BaseEstimator",
private = list(
priv_fields = function(){
classList <- class(self)[-length(class(self))]
classNum <- length(classList)
fieldList <- list()
for (i in 1:classNum) {
priv_fields <- get(class(self)[i])$private_fields
fieldList <- append(fieldList, priv_fields)
}
return(names(fieldList))
},
public_fields = function(){
classList <- class(self)[-length(class(self))]
classNum <- length(classList)
fieldList <- list()
for (i in 1:classNum) {
public_fields <- get(class(self)[i])$public_fields
fieldList <- append(fieldList, public_fields)
}
return(names(fieldList))
}
),
public = list(
#' @description Auxiliary function returning the name of all private and public fields of the self class
#'
#' @examples
#' TestClass <- R6::R6Class(classname = "TestClass",
#' inherit = BaseEstimator,
#' private = list(random_state = NULL))
#' exampleClass <- TestClass$new()
#' exampleClass$get_all_fields()
get_all_fields = function(){
return(append(private$priv_fields(), private$public_fields()))
},
#' @description Auxiliary function returning the name and values of all private and public fields of the self class
#'
#' @examples
#' exampleClass$get_attributes()
get_attributes = function(){
priv_values <- purrr::map(private$priv_fields(), ~.subset2(private, .x))
public_values <- purrr::map(private$public_fields(), ~.subset2(self, .x))
names(priv_values) <- private$priv_fields()
names(public_values) <- private$public_fields()
return(append(priv_values, public_values))
},
#' @description Auxiliary function that sets random state attribute of the self class
#'
#' @param random_state seed number to be set as random state
#' @return self
#'
#' @examples
#' exampleClass$set_random_state(2022)
set_random_state = function(random_state){
private$random_state <- random_state
invisible(self)
}
))
#' @title SklearnEstimator
#'
#' @description A dummy base R6 class that includes fit, predict functions for estimators
#'
#' @return R6 Class of SklearnEstimator
#' @export
SklearnEstimator <- R6::R6Class(classname = "SklearnEstimator",
inherit = BaseEstimator,
private = list(
type = "estimator",
isFitted = FALSE
),
public = list(
#' @description Dummy fit function
#'
#' @examples
#' sklearn <- SklearnEstimator$new()
#' sklearn$fit()
fit = function() {
message("Needs to implement fit(X, y)")
},
#' @description Dummy predict function
#'
#' @examples
#' sklearn$predict()
predict = function(){
message("Needs to implement predict(X, y)")
invisible(self)
},
#' @description Auxiliary function returning the type of the class e.g 'estimator'
#'
#' @examples
#' sklearn$get_type()
get_type = function(){
return(private$type)
},
#' @description Auxiliary function returning the isFitted flag
#'
#' @examples
#' sklearn$get_isFitted()
get_isFitted = function(){
return(private$isFitted)
}
)
)
LocalModel <- R6::R6Class(classname = "LocalModel",
public = list(
estimator = NULL,
center = NULL,
initialize = function(estimator = NULL, center = NULL) {
self$estimator <- estimator
self$center <- center
}
))
Replication <- R6::R6Class(classname = "Replication",
public = list(
sc_object = NULL,
global_estimator = NULL,
local_estimators = NULL,
initialize = function(sc_object = NULL, global_estimator = NULL, local_estimators = NULL) {
self$sc_object <- sc_object
self$global_estimator <- global_estimator
self$local_estimators <- local_estimators
}
))
#' @title LinearRegression
#'
#' @description Wrapper R6 Class of stats::lm function that can be used for LESSRegressor and LESSClassifier
#'
#' @return R6 Class of LinearRegression
#' @importFrom stats lm
#' @seealso [stats::lm()]
#' @export
LinearRegression <- R6::R6Class(classname = "LinearRegression",
inherit = SklearnEstimator,
private = list(
estimator_type = "regressor",
model = NULL
),
public = list(
#' @description Fits a linear model (y ~ X)
#'
#' @param X 2D matrix or dataframe that includes predictors
#' @param y 1D vector or (n,1) dimensional matrix/dataframe that includes response variables
#'
#' @return Fitted R6 Class of LinearRegression
#'
#' @examples
#' data(abalone)
#' split_list <- train_test_split(abalone[1:100,], test_size = 0.3)
#' X_train <- split_list[[1]]
#' X_test <- split_list[[2]]
#' y_train <- split_list[[3]]
#' y_test <- split_list[[4]]
#'
#' lr <- LinearRegression$new()
#' lr$fit(X_train, y_train)
fit = function(X, y) {
df <- prepareDataset(X, y)
private$model <- lm(y ~ ., data = df)
private$isFitted <- TRUE
invisible(self)
},
#' @description Predict regression value for X.
#'
#' @param X0 2D matrix or dataframe that includes predictors
#'
#' @return The predict values.
#'
#' @examples
#' lr <- LinearRegression$new()
#' lr$fit(X_train, y_train)
#' preds <- lr$predict(X_test)
#'
#' lr <- LinearRegression$new()
#' preds <- lr$fit(X_train, y_train)$predict(X_test)
#'
#' preds <- LinearRegression$new()$fit(X_train, y_train)$predict(X_test)
#' print(head(matrix(c(y_test, preds), ncol = 2, dimnames = (list(NULL, c("True", "Prediction"))))))
predict = function(X0) {
check_is_fitted(self)
data <- prepareXset(X0)
# in case of rank deficiency, supress related warnings
suppressWarnings(predict(private$model, newdata = data))
},
#' @description Auxiliary function returning the estimator type e.g 'regressor', 'classifier'
#'
#' @examples
#' lr$get_estimator_type()
get_estimator_type = function() {
return(private$estimator_type)
}
)
)
#' @title DecisionTreeRegressor
#'
#' @description Wrapper R6 Class of rpart::rpart function that can be used for LESSRegressor and LESSClassifier
#'
#' @param min_samples_split The minimum number of observations that must exist in a node in order for a split to be attempted (defaults to 2).
#' @param min_samples_leaf The minimum number of observations in any terminal (leaf) node (defaults to 1).
#' @param cp Complexity Parameter. Any split that does not decrease the overall lack of fit by a factor of cp is not attempted.
#' This means that the overall R-squared must increase by cp at each step. The main role of this parameter is
#' to save computing time by pruning off splits that are obviously not worthwhile. (defaults to 0.001)
#' @param xval Number of cross-validations (defaults to 10)
#' @param surrogate_style Controls the selection of a best surrogate. If set to 0 (default) the program uses the total number of correct
#' classification for a potential surrogate variable, if set to 1 it uses the percent correct, calculated over the non-missing values of the surrogate.
#' The first option more severely penalizes covariates with a large number of missing values.
#' @param max_depth The maximum depth of any node of the final tree, with the root node counted as depth 0.
#' Values greater than 30 will give nonsense results on 32-bit machines.
#'
#' @return R6 Class of DecisionTreeRegressor
#' @importFrom rpart rpart
#' @seealso [rpart::rpart()]
#' @export
DecisionTreeRegressor <- R6::R6Class(classname = "DecisionTreeRegressor",
inherit = SklearnEstimator,
private = list(
estimator_type = "regressor",
model = NULL,
min_samples_split = NULL,
min_samples_leaf = NULL,
cp = NULL,
xval = NULL,
surrogate_style = NULL,
max_depth = NULL
),
public = list(
#' @description Creates a new instance of R6 Class of DecisionTreeRegressor
#'
#' @examples
#' dt <- DecisionTreeRegressor$new()
#' dt <- DecisionTreeRegressor$new(min_samples_split = 10)
#' dt <- DecisionTreeRegressor$new(min_samples_leaf = 6, cp = 0.01)
initialize = function(min_samples_split = 2, min_samples_leaf = 1, cp = 0.001, xval = 10, surrogate_style = 0, max_depth = 30){
private$min_samples_split = min_samples_split
private$min_samples_leaf = min_samples_leaf
private$cp = cp
private$xval = xval
private$surrogate_style = surrogate_style
private$max_depth = max_depth
},
#' @description Builds a decision tree regressor from the training set (X, y).
#'
#' @param X 2D matrix or dataframe that includes predictors
#' @param y 1D vector or (n,1) dimensional matrix/dataframe that includes response variables
#'
#' @return Fitted R6 Class of DecisionTreeRegressor
#'
#' @examples
#' data(abalone)
#' split_list <- train_test_split(abalone[1:100,], test_size = 0.3)
#' X_train <- split_list[[1]]
#' X_test <- split_list[[2]]
#' y_train <- split_list[[3]]
#' y_test <- split_list[[4]]
#'
#' dt <- DecisionTreeRegressor$new()
#' dt$fit(X_train, y_train)
fit = function(X, y) {
df <- prepareDataset(X, y)
private$model <- rpart::rpart(y ~ ., method = "anova", data = df,
control = rpart::rpart.control(minsplit = private$min_samples_split,
minbucket = private$min_samples_leaf,
cp = private$cp, xval = private$xval,
surrogatestyle = private$surrogate_style,
maxdepth = private$max_depth))
private$isFitted <- TRUE
invisible(self)
},
#' @description Predict regression value for X0.
#'
#' @param X0 2D matrix or dataframe that includes predictors
#'
#' @return The predict values.
#'
#' @examples
#' dt <- DecisionTreeRegressor$new()
#' dt$fit(X_train, y_train)
#' preds <- dt$predict(X_test)
#'
#' dt <- DecisionTreeRegressor$new()
#' preds <- dt$fit(X_train, y_train)$predict(X_test)
#'
#' preds <- DecisionTreeRegressor$new()$fit(X_train, y_train)$predict(X_test)
#' print(head(matrix(c(y_test, preds), ncol = 2, dimnames = (list(NULL, c("True", "Prediction"))))))
predict = function(X0) {
check_is_fitted(self)
data <- prepareXset(X0)
predict(private$model, data, method = "anova")
},
#' @description Auxiliary function returning the estimator type e.g 'regressor', 'classifier'
#'
#' @examples
#' dt$get_estimator_type()
get_estimator_type = function() {
return(private$estimator_type)
}
))
#' @title RandomForestRegressor
#'
#' @description Wrapper R6 Class of randomForest::randomForest function that can be used for LESSRegressor and LESSClassifier
#'
#' @param n_estimators Number of trees to grow. This should not be set to too small a number,
#' to ensure that every input row gets predicted at least a few times (defaults to 100).
#' @param random_state Seed number to be used for fixing the randomness (default to NULL).
#' @param min_samples_leaf Minimum size of terminal nodes. Setting this number larger causes smaller trees to be grown
#' (and thus take less time) (defaults to 1)
#' @param max_leaf_nodes Maximum number of terminal nodes trees in the forest can have.
#' If not given, trees are grown to the maximum possible (subject to limits by nodesize).
#' If set larger than maximum possible, a warning is issued. (defaults to NULL)
#'
#' @return R6 Class of RandomForestRegressor
#' @importFrom randomForest randomForest
#' @seealso [randomForest::randomForest()]
#' @export
RandomForestRegressor <- R6::R6Class(classname = "RandomForestRegressor",
inherit = SklearnEstimator,
private = list(
estimator_type = "regressor",
model = NULL,
n_estimators = NULL,
random_state = NULL,
min_samples_leaf = NULL,
max_leaf_nodes = NULL
),
public = list(
#' @description Creates a new instance of R6 Class of RandomForestRegressor
#'
#' @examples
#' rf <- RandomForestRegressor$new()
#' rf <- RandomForestRegressor$new(n_estimators = 500)
#' rf <- RandomForestRegressor$new(n_estimators = 500, random_state = 100)
initialize = function(n_estimators = 100, random_state = NULL, min_samples_leaf = 1,
max_leaf_nodes = NULL){
private$n_estimators = n_estimators
private$random_state = random_state
private$min_samples_leaf = min_samples_leaf
private$max_leaf_nodes = max_leaf_nodes
},
#' @description Builds a random forest regressor from the training set (X, y).
#'
#' @param X 2D matrix or dataframe that includes predictors
#' @param y 1D vector or (n,1) dimensional matrix/dataframe that includes response variables
#'
#' @return Fitted R6 Class of RandomForestRegressor
#'
#' @examples
#' data(abalone)
#' split_list <- train_test_split(abalone[1:100,], test_size = 0.3)
#' X_train <- split_list[[1]]
#' X_test <- split_list[[2]]
#' y_train <- split_list[[3]]
#' y_test <- split_list[[4]]
#'
#' rf <- RandomForestRegressor$new()
#' rf$fit(X_train, y_train)
fit = function(X, y){
df <- prepareDataset(X, y)
set.seed(private$random_state)
private$model <- randomForest::randomForest(y ~ ., data = df, type = "regression", ntree = private$n_estimators,
nodesize = private$min_samples_leaf,
maxnodes = private$max_leaf_nodes)
private$isFitted <- TRUE
invisible(self)
},
#' @description Predict regression value for X0.
#'
#' @param X0 2D matrix or dataframe that includes predictors
#'
#' @return The predict values.
#'
#' @examples
#' preds <- rf$predict(X_test)
#' print(head(matrix(c(y_test, preds), ncol = 2, dimnames = (list(NULL, c("True", "Prediction"))))))
predict = function(X0){
check_is_fitted(self)
data <- prepareXset(X0)
predict(private$model, data)
},
#' @description Auxiliary function returning the estimator type e.g 'regressor', 'classifier'
#'
#' @examples
#' rf$get_estimator_type()
get_estimator_type = function() {
return(private$estimator_type)
}
))
#' @title KNeighborsRegressor
#'
#' @description Wrapper R6 Class of caret::knnreg function that can be used for LESSRegressor and LESSClassifier
#'
#' @param k Number of neighbors considered (defaults to 5).
#'
#' @return R6 Class of KNeighborsRegressor
#' @seealso [caret::knnreg()]
#' @importFrom caret knnreg
#' @export
KNeighborsRegressor <- R6::R6Class(classname = "KNeighborsRegressor",
inherit = SklearnEstimator,
private = list(
estimator_type = "regressor",
model = NULL,
k = NULL
),
public = list(
#' @description Creates a new instance of R6 Class of KNeighborsRegressor
#'
#' @examples
#' knr <- KNeighborsRegressor$new()
#' knr <- KNeighborsRegressor$new(k = 5)
initialize = function(k = 5){
private$k = k
},
#' @description Fit the k-nearest neighbors regressor from the training set (X, y).
#'
#' @param X 2D matrix or dataframe that includes predictors
#' @param y 1D vector or (n,1) dimensional matrix/dataframe that includes response variables
#'
#' @return Fitted R6 Class of KNeighborsRegressor
#'
#' @examples
#' data(abalone)
#' split_list <- train_test_split(abalone[1:100,], test_size = 0.3)
#' X_train <- split_list[[1]]
#' X_test <- split_list[[2]]
#' y_train <- split_list[[3]]
#' y_test <- split_list[[4]]
#'
#' knr <- KNeighborsRegressor$new()
#' knr$fit(X_train, y_train)
fit = function(X, y){
df <- prepareDataset(X, y)
private$model <- caret::knnreg(y ~ ., data = df, k=private$k)
private$isFitted <- TRUE
invisible(self)
},
#' @description Predict regression value for X0.
#'
#' @param X0 2D matrix or dataframe that includes predictors
#'
#' @return The predict values.
#'
#' @examples
#' knr <- KNeighborsRegressor$new()
#' knr$fit(X_train, y_train)
#' preds <- knr$predict(X_test)
#'
#' knr <- KNeighborsRegressor$new()
#' preds <- knr$fit(X_train, y_train)$predict(X_test)
#'
#' preds <- KNeighborsRegressor$new()$fit(X_train, y_train)$predict(X_test)
#' print(head(matrix(c(y_test, preds), ncol = 2, dimnames = (list(NULL, c("True", "Prediction"))))))
predict = function(X0){
check_is_fitted(self)
data <- prepareXset(X0)
predict(private$model, data)
},
#' @description Auxiliary function returning the estimator type e.g 'regressor', 'classifier'
#'
#' @examples
#' knr$get_estimator_type()
get_estimator_type = function() {
return(private$estimator_type)
}
))
#' @title Support Vector Regression
#'
#' @description Wrapper R6 Class of e1071::svm function that can be used for LESSRegressor and LESSClassifier
#'
#' @param scale A logical vector indicating the variables to be scaled. If scale is of length 1, the value is recycled as many times as needed.
#' Per default, data are scaled internally (both x and y variables) to zero mean and unit variance.
#' The center and scale values are returned and used for later predictions (default: TRUE)
#' @param kernel The kernel used in training and predicting. Possible values are: "linear", "polynomial", "radial", "sigmoid" (default is "radial")
#' @param degree Parameter needed for kernel of type polynomial (default: 3)
#' @param gamma Parameter needed for all kernels except linear (default: 1/(data dimension))
#' @param cost Cost of constraints violation (default: 1)—it is the ‘C’-constant of the regularization term in the Lagrange formulation (default: 1)
#' @param cache_size Cache memory in MB (default: 40)
#' @param coef0 Parameter needed for kernels of type polynomial and sigmoid (default: 0)
#' @param tolerance Tolerance of termination criterion (default: 0.001)
#' @param epsilon Epsilon in the insensitive-loss function (default: 0.1)
#' @param shrinking Option whether to use the shrinking-heuristics (default: TRUE)
#' @param cross If a integer value k>0 is specified, a k-fold cross validation on the training data is performed to assess the quality of the model:
#' the accuracy rate for classification and the Mean Squared Error for regression (default: 0)
#' @param fitted Logical indicating whether the fitted values should be computed and included in the model or not (default: TRUE)
#' @param probability Logical indicating whether the model should allow for probability predictions (default: FALSE)
#'
#' @return R6 Class of SVR
#' @seealso [e1071::svm()]
#' @importFrom e1071 svm
#' @export
SVR <- R6::R6Class(classname = "SVR",
inherit = SklearnEstimator,
private = list(
estimator_type = "regressor",
model = NULL,
scale = NULL,
kernel = NULL,
degree = NULL,
gamma = NULL,
coef0 = NULL,
cost = NULL,
cache_size = NULL,
tolerance = NULL,
epsilon = NULL,
shrinking = NULL,
cross = NULL,
probability = NULL,
fitted = NULL
),
public = list(
#' @description Creates a new instance of R6 Class of SVR
#'
#' @examples
#' svr <- SVR$new()
#' svr <- SVR$new(kernel = "polynomial")
initialize = function(scale = TRUE, kernel = "radial", degree = 3, gamma = NULL, coef0 = 0, cost = 1,
cache_size = 40, tolerance = 0.001, epsilon = 0.1, shrinking = TRUE, cross = 0,
probability = FALSE, fitted = TRUE){
private$scale = scale
private$kernel = kernel
private$degree = degree
private$gamma = gamma
private$coef0 = coef0
private$cost = cost
private$cache_size = cache_size
private$tolerance = tolerance
private$epsilon = epsilon
private$shrinking = shrinking
private$cross = cross
private$probability = probability
private$fitted = fitted
},
#' @description Fit the SVM model from the training set (X, y).
#'
#' @param X 2D matrix or dataframe that includes predictors
#' @param y 1D vector or (n,1) dimensional matrix/dataframe that includes response variables
#'
#' @return Fitted R6 Class of SVR
#'
#' @examples
#' data(abalone)
#' split_list <- train_test_split(abalone[1:100,], test_size = 0.3)
#' X_train <- split_list[[1]]
#' X_test <- split_list[[2]]
#' y_train <- split_list[[3]]
#' y_test <- split_list[[4]]
#'
#' svr <- SVR$new()
#' svr$fit(X_train, y_train)
fit = function(X, y){
df <- prepareDataset(X, y)
gamma = if (is.vector(X)) 1 else 1 / ncol(X)
private$model <- e1071::svm(y ~ ., data = df, scale = private$scale, kernel = private$kernel, degree = private$degree,
gamma = if (is.null(private$gamma)) gamma else private$gamma, coef0 = private$coef0,
cost = private$cost, cachesize = private$cache_size, tolerance = private$tolerance, epsilon = private$epsilon,
shrinking = private$shrinking, cross = private$cross, probability = private$probability,
fitted = private$fitted)
private$isFitted <- TRUE
invisible(self)
},
#' @description Predict regression value for X0.
#'
#' @param X0 2D matrix or dataframe that includes predictors
#'
#' @return The predict values.
#'
#' @examples
#' svr <- SVR$new()
#' svr$fit(X_train, y_train)
#' preds <- svr$predict(X_test)
#'
#' svr <- SVR$new()
#' preds <- svr$fit(X_train, y_train)$predict(X_test)
#'
#' preds <- SVR$new()$fit(X_train, y_train)$predict(X_test)
#' print(head(matrix(c(y_test, preds), ncol = 2, dimnames = (list(NULL, c("True", "Prediction"))))))
predict = function(X0){
check_is_fitted(self)
data <- prepareXset(X0)
predict(private$model, data)
},
#' @description Auxiliary function returning the estimator type e.g 'regressor', 'classifier'
#'
#' @examples
#' svr$get_estimator_type()
get_estimator_type = function() {
return(private$estimator_type)
}
))
Try the less package in your browser
Any scripts or data that you put into this service are public.
less documentation built on Sept. 27, 2022, 5:05 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
####################
# HELPER CLASSES
####################
#' @title BaseEstimator
#'
#' @description A dummy base R6 class that provides get_all_fields, get_attributes and set_random_state functionalities for estimators
#'
#' @return R6 Class of BaseEstimator
#' @export
BaseEstimator <- R6::R6Class(classname = "BaseEstimator",
private = list(
priv_fields = function(){
classList <- class(self)[-length(class(self))]
classNum <- length(classList)
fieldList <- list()
for (i in 1:classNum) {
priv_fields <- get(class(self)[i])$private_fields
fieldList <- append(fieldList, priv_fields)
}
return(names(fieldList))
},
public_fields = function(){
classList <- class(self)[-length(class(self))]
classNum <- length(classList)
fieldList <- list()
for (i in 1:classNum) {
public_fields <- get(class(self)[i])$public_fields
fieldList <- append(fieldList, public_fields)
}
return(names(fieldList))
}
),
public = list(
#' @description Auxiliary function returning the name of all private and public fields of the self class
#'
#' @examples
#' TestClass <- R6::R6Class(classname = "TestClass",
#' inherit = BaseEstimator,
#' private = list(random_state = NULL))
#' exampleClass <- TestClass$new()
#' exampleClass$get_all_fields()
get_all_fields = function(){
return(append(private$priv_fields(), private$public_fields()))
},
#' @description Auxiliary function returning the name and values of all private and public fields of the self class
#'
#' @examples
#' exampleClass$get_attributes()
get_attributes = function(){
priv_values <- purrr::map(private$priv_fields(), ~.subset2(private, .x))
public_values <- purrr::map(private$public_fields(), ~.subset2(self, .x))
names(priv_values) <- private$priv_fields()
names(public_values) <- private$public_fields()
return(append(priv_values, public_values))
},
#' @description Auxiliary function that sets random state attribute of the self class
#'
#' @param random_state seed number to be set as random state
#' @return self
#'
#' @examples
#' exampleClass$set_random_state(2022)
set_random_state = function(random_state){
private$random_state <- random_state
invisible(self)
}
))
#' @title SklearnEstimator
#'
#' @description A dummy base R6 class that includes fit, predict functions for estimators
#'
#' @return R6 Class of SklearnEstimator
#' @export
SklearnEstimator <- R6::R6Class(classname = "SklearnEstimator",
inherit = BaseEstimator,
private = list(
type = "estimator",
isFitted = FALSE
),
public = list(
#' @description Dummy fit function
#'
#' @examples
#' sklearn <- SklearnEstimator$new()
#' sklearn$fit()
fit = function() {
message("Needs to implement fit(X, y)")
},
#' @description Dummy predict function
#'
#' @examples
#' sklearn$predict()
predict = function(){
message("Needs to implement predict(X, y)")
invisible(self)
},
#' @description Auxiliary function returning the type of the class e.g 'estimator'
#'
#' @examples
#' sklearn$get_type()
get_type = function(){
return(private$type)
},
#' @description Auxiliary function returning the isFitted flag
#'
#' @examples
#' sklearn$get_isFitted()
get_isFitted = function(){
return(private$isFitted)
}
)
)
LocalModel <- R6::R6Class(classname = "LocalModel",
public = list(
estimator = NULL,
center = NULL,
initialize = function(estimator = NULL, center = NULL) {
self$estimator <- estimator
self$center <- center
}
))
Replication <- R6::R6Class(classname = "Replication",
public = list(
sc_object = NULL,
global_estimator = NULL,
local_estimators = NULL,
initialize = function(sc_object = NULL, global_estimator = NULL, local_estimators = NULL) {
self$sc_object <- sc_object
self$global_estimator <- global_estimator
self$local_estimators <- local_estimators
}
))
#' @title LinearRegression
#'
#' @description Wrapper R6 Class of stats::lm function that can be used for LESSRegressor and LESSClassifier
#'
#' @return R6 Class of LinearRegression
#' @importFrom stats lm
#' @seealso [stats::lm()]
#' @export
LinearRegression <- R6::R6Class(classname = "LinearRegression",
inherit = SklearnEstimator,
private = list(
estimator_type = "regressor",
model = NULL
),
public = list(
#' @description Fits a linear model (y ~ X)
#'
#' @param X 2D matrix or dataframe that includes predictors
#' @param y 1D vector or (n,1) dimensional matrix/dataframe that includes response variables
#'
#' @return Fitted R6 Class of LinearRegression
#'
#' @examples
#' data(abalone)
#' split_list <- train_test_split(abalone[1:100,], test_size = 0.3)
#' X_train <- split_list[[1]]
#' X_test <- split_list[[2]]
#' y_train <- split_list[[3]]
#' y_test <- split_list[[4]]
#'
#' lr <- LinearRegression$new()
#' lr$fit(X_train, y_train)
fit = function(X, y) {
df <- prepareDataset(X, y)
private$model <- lm(y ~ ., data = df)
private$isFitted <- TRUE
invisible(self)
},
#' @description Predict regression value for X.
#'
#' @param X0 2D matrix or dataframe that includes predictors
#'
#' @return The predict values.
#'
#' @examples
#' lr <- LinearRegression$new()
#' lr$fit(X_train, y_train)
#' preds <- lr$predict(X_test)
#'
#' lr <- LinearRegression$new()
#' preds <- lr$fit(X_train, y_train)$predict(X_test)
#'
#' preds <- LinearRegression$new()$fit(X_train, y_train)$predict(X_test)
#' print(head(matrix(c(y_test, preds), ncol = 2, dimnames = (list(NULL, c("True", "Prediction"))))))
predict = function(X0) {
check_is_fitted(self)
data <- prepareXset(X0)
# in case of rank deficiency, supress related warnings
suppressWarnings(predict(private$model, newdata = data))
},
#' @description Auxiliary function returning the estimator type e.g 'regressor', 'classifier'
#'
#' @examples
#' lr$get_estimator_type()
get_estimator_type = function() {
return(private$estimator_type)
}
)
)
#' @title DecisionTreeRegressor
#'
#' @description Wrapper R6 Class of rpart::rpart function that can be used for LESSRegressor and LESSClassifier
#'
#' @param min_samples_split The minimum number of observations that must exist in a node in order for a split to be attempted (defaults to 2).
#' @param min_samples_leaf The minimum number of observations in any terminal (leaf) node (defaults to 1).
#' @param cp Complexity Parameter. Any split that does not decrease the overall lack of fit by a factor of cp is not attempted.
#' This means that the overall R-squared must increase by cp at each step. The main role of this parameter is
#' to save computing time by pruning off splits that are obviously not worthwhile. (defaults to 0.001)
#' @param xval Number of cross-validations (defaults to 10)
#' @param surrogate_style Controls the selection of a best surrogate. If set to 0 (default) the program uses the total number of correct
#' classification for a potential surrogate variable, if set to 1 it uses the percent correct, calculated over the non-missing values of the surrogate.
#' The first option more severely penalizes covariates with a large number of missing values.
#' @param max_depth The maximum depth of any node of the final tree, with the root node counted as depth 0.
#' Values greater than 30 will give nonsense results on 32-bit machines.
#'
#' @return R6 Class of DecisionTreeRegressor
#' @importFrom rpart rpart
#' @seealso [rpart::rpart()]
#' @export
DecisionTreeRegressor <- R6::R6Class(classname = "DecisionTreeRegressor",
inherit = SklearnEstimator,
private = list(
estimator_type = "regressor",
model = NULL,
min_samples_split = NULL,
min_samples_leaf = NULL,
cp = NULL,
xval = NULL,
surrogate_style = NULL,
max_depth = NULL
),
public = list(
#' @description Creates a new instance of R6 Class of DecisionTreeRegressor
#'
#' @examples
#' dt <- DecisionTreeRegressor$new()
#' dt <- DecisionTreeRegressor$new(min_samples_split = 10)
#' dt <- DecisionTreeRegressor$new(min_samples_leaf = 6, cp = 0.01)
initialize = function(min_samples_split = 2, min_samples_leaf = 1, cp = 0.001, xval = 10, surrogate_style = 0, max_depth = 30){
private$min_samples_split = min_samples_split
private$min_samples_leaf = min_samples_leaf
private$cp = cp
private$xval = xval
private$surrogate_style = surrogate_style
private$max_depth = max_depth
},
#' @description Builds a decision tree regressor from the training set (X, y).
#'
#' @param X 2D matrix or dataframe that includes predictors
#' @param y 1D vector or (n,1) dimensional matrix/dataframe that includes response variables
#'
#' @return Fitted R6 Class of DecisionTreeRegressor
#'
#' @examples
#' data(abalone)
#' split_list <- train_test_split(abalone[1:100,], test_size = 0.3)
#' X_train <- split_list[[1]]
#' X_test <- split_list[[2]]
#' y_train <- split_list[[3]]
#' y_test <- split_list[[4]]
#'
#' dt <- DecisionTreeRegressor$new()
#' dt$fit(X_train, y_train)
fit = function(X, y) {
df <- prepareDataset(X, y)
private$model <- rpart::rpart(y ~ ., method = "anova", data = df,
control = rpart::rpart.control(minsplit = private$min_samples_split,
minbucket = private$min_samples_leaf,
cp = private$cp, xval = private$xval,
surrogatestyle = private$surrogate_style,
maxdepth = private$max_depth))
private$isFitted <- TRUE
invisible(self)
},
#' @description Predict regression value for X0.
#'
#' @param X0 2D matrix or dataframe that includes predictors
#'
#' @return The predict values.
#'
#' @examples
#' dt <- DecisionTreeRegressor$new()
#' dt$fit(X_train, y_train)
#' preds <- dt$predict(X_test)
#'
#' dt <- DecisionTreeRegressor$new()
#' preds <- dt$fit(X_train, y_train)$predict(X_test)
#'
#' preds <- DecisionTreeRegressor$new()$fit(X_train, y_train)$predict(X_test)
#' print(head(matrix(c(y_test, preds), ncol = 2, dimnames = (list(NULL, c("True", "Prediction"))))))
predict = function(X0) {
check_is_fitted(self)
data <- prepareXset(X0)
predict(private$model, data, method = "anova")
},
#' @description Auxiliary function returning the estimator type e.g 'regressor', 'classifier'
#'
#' @examples
#' dt$get_estimator_type()
get_estimator_type = function() {
return(private$estimator_type)
}
))
#' @title RandomForestRegressor
#'
#' @description Wrapper R6 Class of randomForest::randomForest function that can be used for LESSRegressor and LESSClassifier
#'
#' @param n_estimators Number of trees to grow. This should not be set to too small a number,
#' to ensure that every input row gets predicted at least a few times (defaults to 100).
#' @param random_state Seed number to be used for fixing the randomness (default to NULL).
#' @param min_samples_leaf Minimum size of terminal nodes. Setting this number larger causes smaller trees to be grown
#' (and thus take less time) (defaults to 1)
#' @param max_leaf_nodes Maximum number of terminal nodes trees in the forest can have.
#' If not given, trees are grown to the maximum possible (subject to limits by nodesize).
#' If set larger than maximum possible, a warning is issued. (defaults to NULL)
#'
#' @return R6 Class of RandomForestRegressor
#' @importFrom randomForest randomForest
#' @seealso [randomForest::randomForest()]
#' @export
RandomForestRegressor <- R6::R6Class(classname = "RandomForestRegressor",
inherit = SklearnEstimator,
private = list(
estimator_type = "regressor",
model = NULL,
n_estimators = NULL,
random_state = NULL,
min_samples_leaf = NULL,
max_leaf_nodes = NULL
),
public = list(
#' @description Creates a new instance of R6 Class of RandomForestRegressor
#'
#' @examples
#' rf <- RandomForestRegressor$new()
#' rf <- RandomForestRegressor$new(n_estimators = 500)
#' rf <- RandomForestRegressor$new(n_estimators = 500, random_state = 100)
initialize = function(n_estimators = 100, random_state = NULL, min_samples_leaf = 1,
max_leaf_nodes = NULL){
private$n_estimators = n_estimators
private$random_state = random_state
private$min_samples_leaf = min_samples_leaf
private$max_leaf_nodes = max_leaf_nodes
},
#' @description Builds a random forest regressor from the training set (X, y).
#'
#' @param X 2D matrix or dataframe that includes predictors
#' @param y 1D vector or (n,1) dimensional matrix/dataframe that includes response variables
#'
#' @return Fitted R6 Class of RandomForestRegressor
#'
#' @examples
#' data(abalone)
#' split_list <- train_test_split(abalone[1:100,], test_size = 0.3)
#' X_train <- split_list[[1]]
#' X_test <- split_list[[2]]
#' y_train <- split_list[[3]]
#' y_test <- split_list[[4]]
#'
#' rf <- RandomForestRegressor$new()
#' rf$fit(X_train, y_train)
fit = function(X, y){
df <- prepareDataset(X, y)
set.seed(private$random_state)
private$model <- randomForest::randomForest(y ~ ., data = df, type = "regression", ntree = private$n_estimators,
nodesize = private$min_samples_leaf,
maxnodes = private$max_leaf_nodes)
private$isFitted <- TRUE
invisible(self)
},
#' @description Predict regression value for X0.
#'
#' @param X0 2D matrix or dataframe that includes predictors
#'
#' @return The predict values.
#'
#' @examples
#' preds <- rf$predict(X_test)
#' print(head(matrix(c(y_test, preds), ncol = 2, dimnames = (list(NULL, c("True", "Prediction"))))))
predict = function(X0){
check_is_fitted(self)
data <- prepareXset(X0)
predict(private$model, data)
},
#' @description Auxiliary function returning the estimator type e.g 'regressor', 'classifier'
#'
#' @examples
#' rf$get_estimator_type()
get_estimator_type = function() {
return(private$estimator_type)
}
))
#' @title KNeighborsRegressor
#'
#' @description Wrapper R6 Class of caret::knnreg function that can be used for LESSRegressor and LESSClassifier
#'
#' @param k Number of neighbors considered (defaults to 5).
#'
#' @return R6 Class of KNeighborsRegressor
#' @seealso [caret::knnreg()]
#' @importFrom caret knnreg
#' @export
KNeighborsRegressor <- R6::R6Class(classname = "KNeighborsRegressor",
inherit = SklearnEstimator,
private = list(
estimator_type = "regressor",
model = NULL,
k = NULL
),
public = list(
#' @description Creates a new instance of R6 Class of KNeighborsRegressor
#'
#' @examples
#' knr <- KNeighborsRegressor$new()
#' knr <- KNeighborsRegressor$new(k = 5)
initialize = function(k = 5){
private$k = k
},
#' @description Fit the k-nearest neighbors regressor from the training set (X, y).
#'
#' @param X 2D matrix or dataframe that includes predictors
#' @param y 1D vector or (n,1) dimensional matrix/dataframe that includes response variables
#'
#' @return Fitted R6 Class of KNeighborsRegressor
#'
#' @examples
#' data(abalone)
#' split_list <- train_test_split(abalone[1:100,], test_size = 0.3)
#' X_train <- split_list[[1]]
#' X_test <- split_list[[2]]
#' y_train <- split_list[[3]]
#' y_test <- split_list[[4]]
#'
#' knr <- KNeighborsRegressor$new()
#' knr$fit(X_train, y_train)
fit = function(X, y){
df <- prepareDataset(X, y)
private$model <- caret::knnreg(y ~ ., data = df, k=private$k)
private$isFitted <- TRUE
invisible(self)
},
#' @description Predict regression value for X0.
#'
#' @param X0 2D matrix or dataframe that includes predictors
#'
#' @return The predict values.
#'
#' @examples
#' knr <- KNeighborsRegressor$new()
#' knr$fit(X_train, y_train)
#' preds <- knr$predict(X_test)
#'
#' knr <- KNeighborsRegressor$new()
#' preds <- knr$fit(X_train, y_train)$predict(X_test)
#'
#' preds <- KNeighborsRegressor$new()$fit(X_train, y_train)$predict(X_test)
#' print(head(matrix(c(y_test, preds), ncol = 2, dimnames = (list(NULL, c("True", "Prediction"))))))
predict = function(X0){
check_is_fitted(self)
data <- prepareXset(X0)
predict(private$model, data)
},
#' @description Auxiliary function returning the estimator type e.g 'regressor', 'classifier'
#'
#' @examples
#' knr$get_estimator_type()
get_estimator_type = function() {
return(private$estimator_type)
}
))
#' @title Support Vector Regression
#'
#' @description Wrapper R6 Class of e1071::svm function that can be used for LESSRegressor and LESSClassifier
#'
#' @param scale A logical vector indicating the variables to be scaled. If scale is of length 1, the value is recycled as many times as needed.
#' Per default, data are scaled internally (both x and y variables) to zero mean and unit variance.
#' The center and scale values are returned and used for later predictions (default: TRUE)
#' @param kernel The kernel used in training and predicting. Possible values are: "linear", "polynomial", "radial", "sigmoid" (default is "radial")
#' @param degree Parameter needed for kernel of type polynomial (default: 3)
#' @param gamma Parameter needed for all kernels except linear (default: 1/(data dimension))
#' @param cost Cost of constraints violation (default: 1)—it is the ‘C’-constant of the regularization term in the Lagrange formulation (default: 1)
#' @param cache_size Cache memory in MB (default: 40)
#' @param coef0 Parameter needed for kernels of type polynomial and sigmoid (default: 0)
#' @param tolerance Tolerance of termination criterion (default: 0.001)
#' @param epsilon Epsilon in the insensitive-loss function (default: 0.1)
#' @param shrinking Option whether to use the shrinking-heuristics (default: TRUE)
#' @param cross If a integer value k>0 is specified, a k-fold cross validation on the training data is performed to assess the quality of the model:
#' the accuracy rate for classification and the Mean Squared Error for regression (default: 0)
#' @param fitted Logical indicating whether the fitted values should be computed and included in the model or not (default: TRUE)
#' @param probability Logical indicating whether the model should allow for probability predictions (default: FALSE)
#'
#' @return R6 Class of SVR
#' @seealso [e1071::svm()]
#' @importFrom e1071 svm
#' @export
SVR <- R6::R6Class(classname = "SVR",
inherit = SklearnEstimator,
private = list(
estimator_type = "regressor",
model = NULL,
scale = NULL,
kernel = NULL,
degree = NULL,
gamma = NULL,
coef0 = NULL,
cost = NULL,
cache_size = NULL,
tolerance = NULL,
epsilon = NULL,
shrinking = NULL,
cross = NULL,
probability = NULL,
fitted = NULL
),
public = list(
#' @description Creates a new instance of R6 Class of SVR
#'
#' @examples
#' svr <- SVR$new()
#' svr <- SVR$new(kernel = "polynomial")
initialize = function(scale = TRUE, kernel = "radial", degree = 3, gamma = NULL, coef0 = 0, cost = 1,
cache_size = 40, tolerance = 0.001, epsilon = 0.1, shrinking = TRUE, cross = 0,
probability = FALSE, fitted = TRUE){
private$scale = scale
private$kernel = kernel
private$degree = degree
private$gamma = gamma
private$coef0 = coef0
private$cost = cost
private$cache_size = cache_size
private$tolerance = tolerance
private$epsilon = epsilon
private$shrinking = shrinking
private$cross = cross
private$probability = probability
private$fitted = fitted
},
#' @description Fit the SVM model from the training set (X, y).
#'
#' @param X 2D matrix or dataframe that includes predictors
#' @param y 1D vector or (n,1) dimensional matrix/dataframe that includes response variables
#'
#' @return Fitted R6 Class of SVR
#'
#' @examples
#' data(abalone)
#' split_list <- train_test_split(abalone[1:100,], test_size = 0.3)
#' X_train <- split_list[[1]]
#' X_test <- split_list[[2]]
#' y_train <- split_list[[3]]
#' y_test <- split_list[[4]]
#'
#' svr <- SVR$new()
#' svr$fit(X_train, y_train)
fit = function(X, y){
df <- prepareDataset(X, y)
gamma = if (is.vector(X)) 1 else 1 / ncol(X)
private$model <- e1071::svm(y ~ ., data = df, scale = private$scale, kernel = private$kernel, degree = private$degree,
gamma = if (is.null(private$gamma)) gamma else private$gamma, coef0 = private$coef0,
cost = private$cost, cachesize = private$cache_size, tolerance = private$tolerance, epsilon = private$epsilon,
shrinking = private$shrinking, cross = private$cross, probability = private$probability,
fitted = private$fitted)
private$isFitted <- TRUE
invisible(self)
},
#' @description Predict regression value for X0.
#'
#' @param X0 2D matrix or dataframe that includes predictors
#'
#' @return The predict values.
#'
#' @examples
#' svr <- SVR$new()
#' svr$fit(X_train, y_train)
#' preds <- svr$predict(X_test)
#'
#' svr <- SVR$new()
#' preds <- svr$fit(X_train, y_train)$predict(X_test)
#'
#' preds <- SVR$new()$fit(X_train, y_train)$predict(X_test)
#' print(head(matrix(c(y_test, preds), ncol = 2, dimnames = (list(NULL, c("True", "Prediction"))))))
predict = function(X0){
check_is_fitted(self)
data <- prepareXset(X0)
predict(private$model, data)
},
#' @description Auxiliary function returning the estimator type e.g 'regressor', 'classifier'
#'
#' @examples
#' svr$get_estimator_type()
get_estimator_type = function() {
return(private$estimator_type)
}
))
Try the less package in your browser
Any scripts or data that you put into this service are public.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.