Hendrik/xgboost
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

rename arguments to be dot-seperated

Browse Source
This commit is contained in:
hetong007 2015-05-25 11:51:01 -07:00
parent 8d3a7e1688
commit 733d23aef8
6 changed files with 40 additions and 55 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
R-package/R/xgb.cv.R
View File

@ -54,12 +54,11 @@
#' @param folds \code{list} provides a possibility of using a list of pre-defined CV folds (each element must be a vector of fold's indices). #' @param folds \code{list} provides a possibility of using a list of pre-defined CV folds (each element must be a vector of fold's indices).
#' If folds are supplied, the nfold and stratified parameters would be ignored. #' If folds are supplied, the nfold and stratified parameters would be ignored.
#' @param verbose \code{boolean}, print the statistics during the process #' @param verbose \code{boolean}, print the statistics during the process
#' @param printEveryN Print every N progress messages when \code{verbose>0}. Default is 1 which means all messages are printed. #' @param print.every.n Print every N progress messages when \code{verbose>0}. Default is 1 which means all messages are printed.
#' @param early_stop_round If \code{NULL}, the early stopping function is not triggered. #' @param early.stop.round If \code{NULL}, the early stopping function is not triggered.
#' If set to an integer \code{k}, training with a validation set will stop if the performance #' If set to an integer \code{k}, training with a validation set will stop if the performance
#' keeps getting worse consecutively for \code{k} rounds. #' keeps getting worse consecutively for \code{k} rounds.
#' @param early.stop.round An alternative of \code{early_stop_round}. #' @param maximize If \code{feval} and \code{early.stop.round} are set, then \code{maximize} must be set as well.
#' @param maximize If \code{feval} and \code{early_stop_round} are set, then \code{maximize} must be set as well.
#' \code{maximize=TRUE} means the larger the evaluation score the better. #' \code{maximize=TRUE} means the larger the evaluation score the better.
#' #'
#' @param ... other parameters to pass to \code{params}. #' @param ... other parameters to pass to \code{params}.
@ -94,8 +93,8 @@
#' #'
xgb.cv <- function(params=list(), data, nrounds, nfold, label = NULL, missing = NULL, xgb.cv <- function(params=list(), data, nrounds, nfold, label = NULL, missing = NULL,
prediction = FALSE, showsd = TRUE, metrics=list(), prediction = FALSE, showsd = TRUE, metrics=list(),
obj = NULL, feval = NULL, stratified = TRUE, folds = NULL, verbose = T, printEveryN=1L, obj = NULL, feval = NULL, stratified = TRUE, folds = NULL, verbose = T, print.every.n=1L,
early_stop_round = NULL, early.stop.round = NULL, maximize = NULL, ...) { early.stop.round = NULL, maximize = NULL, ...) {
if (typeof(params) != "list") { if (typeof(params) != "list") {
stop("xgb.cv: first argument params must be list") stop("xgb.cv: first argument params must be list")
} }
@ -136,9 +135,7 @@ xgb.cv <- function(params=list(), data, nrounds, nfold, label = NULL, missing =
} }
# Early Stopping # Early Stopping
if (is.null(early_stop_round) && !is.null(early.stop.round)) if (!is.null(early.stop.round)){
early_stop_round = early.stop.round
if (!is.null(early_stop_round)){
if (!is.null(feval) && is.null(maximize)) if (!is.null(feval) && is.null(maximize))
stop('Please set maximize to note whether the model is maximizing the evaluation or not.') stop('Please set maximize to note whether the model is maximizing the evaluation or not.')
if (is.null(maximize) && is.null(params$eval_metric)) if (is.null(maximize) && is.null(params$eval_metric))
@ -178,7 +175,7 @@ xgb.cv <- function(params=list(), data, nrounds, nfold, label = NULL, missing =
else else
predictValues <- rep(0,xgb.numrow(dtrain)) predictValues <- rep(0,xgb.numrow(dtrain))
history <- c() history <- c()
printEveryN = max(as.integer(printEveryN), 1L) print.every.n = max(as.integer(print.every.n), 1L)
for (i in 1:nrounds) { for (i in 1:nrounds) {
msg <- list() msg <- list()
for (k in 1:nfold) { for (k in 1:nfold) {
@ -204,11 +201,11 @@ xgb.cv <- function(params=list(), data, nrounds, nfold, label = NULL, missing =
ret <- xgb.cv.aggcv(msg, showsd) ret <- xgb.cv.aggcv(msg, showsd)
history <- c(history, ret) history <- c(history, ret)
if(verbose) if(verbose)
if (0==(i-1L)%%printEveryN) if (0==(i-1L)%%print.every.n)
cat(ret, "\n", sep="") cat(ret, "\n", sep="")
# early_Stopping # early_Stopping
if (!is.null(early_stop_round)){ if (!is.null(early.stop.round)){
score = strsplit(ret,'\\s+')[[1]][1+length(metrics)+1] score = strsplit(ret,'\\s+')[[1]][1+length(metrics)+1]
score = strsplit(score,'\\+|:')[[1]][[2]] score = strsplit(score,'\\+|:')[[1]][[2]]
score = as.numeric(score) score = as.numeric(score)
@ -216,7 +213,7 @@ xgb.cv <- function(params=list(), data, nrounds, nfold, label = NULL, missing =
bestScore = score bestScore = score
bestInd = i bestInd = i
} else { } else {
if (i-bestInd>=early_stop_round) { if (i-bestInd>=early.stop.round) {
earlyStopflag = TRUE earlyStopflag = TRUE
cat('Stopping. Best iteration:',bestInd) cat('Stopping. Best iteration:',bestInd)
break break

26
R-package/R/xgb.train.R
View File

@ -66,12 +66,11 @@
#' prediction and dtrain, #' prediction and dtrain,
#' @param verbose If 0, xgboost will stay silent. If 1, xgboost will print #' @param verbose If 0, xgboost will stay silent. If 1, xgboost will print
#' information of performance. If 2, xgboost will print information of both #' information of performance. If 2, xgboost will print information of both
#' @param printEveryN Print every N progress messages when \code{verbose>0}. Default is 1 which means all messages are printed. #' @param print.every.n Print every N progress messages when \code{verbose>0}. Default is 1 which means all messages are printed.
#' @param early_stop_round If \code{NULL}, the early stopping function is not triggered. #' @param early.stop.round If \code{NULL}, the early stopping function is not triggered.
#' If set to an integer \code{k}, training with a validation set will stop if the performance #' If set to an integer \code{k}, training with a validation set will stop if the performance
#' keeps getting worse consecutively for \code{k} rounds. #' keeps getting worse consecutively for \code{k} rounds.
#' @param early.stop.round An alternative of \code{early_stop_round}. #' @param maximize If \code{feval} and \code{early.stop.round} are set, then \code{maximize} must be set as well.
#' @param maximize If \code{feval} and \code{early_stop_round} are set, then \code{maximize} must be set as well.
#' \code{maximize=TRUE} means the larger the evaluation score the better. #' \code{maximize=TRUE} means the larger the evaluation score the better.
#' @param ... other parameters to pass to \code{params}. #' @param ... other parameters to pass to \code{params}.
#' #'
@ -120,9 +119,8 @@
#' @export #' @export
#' #'
xgb.train <- function(params=list(), data, nrounds, watchlist = list(), xgb.train <- function(params=list(), data, nrounds, watchlist = list(),
obj = NULL, feval = NULL, verbose = 1, printEveryN=1L, obj = NULL, feval = NULL, verbose = 1, print.every.n=1L,
early_stop_round = NULL, early.stop.round = NULL, early.stop.round = NULL, maximize = NULL, ...) {
maximize = NULL, ...) {
dtrain <- data dtrain <- data
if (typeof(params) != "list") { if (typeof(params) != "list") {
stop("xgb.train: first argument params must be list") stop("xgb.train: first argument params must be list")
@ -157,9 +155,7 @@ xgb.train <- function(params=list(), data, nrounds, watchlist = list(),
} }
# Early stopping # Early stopping
if (is.null(early_stop_round) && !is.null(early.stop.round)) if (!is.null(early.stop.round)){
early_stop_round = early.stop.round
if (!is.null(early_stop_round)){
if (!is.null(feval) && is.null(maximize)) if (!is.null(feval) && is.null(maximize))
stop('Please set maximize to note whether the model is maximizing the evaluation or not.') stop('Please set maximize to note whether the model is maximizing the evaluation or not.')
if (length(watchlist) == 0) if (length(watchlist) == 0)
@ -190,14 +186,14 @@ xgb.train <- function(params=list(), data, nrounds, watchlist = list(),
handle <- xgb.Booster(params, append(watchlist, dtrain)) handle <- xgb.Booster(params, append(watchlist, dtrain))
bst <- xgb.handleToBooster(handle) bst <- xgb.handleToBooster(handle)
printEveryN=max( as.integer(printEveryN), 1L) print.every.n=max( as.integer(print.every.n), 1L)
for (i in 1:nrounds) { for (i in 1:nrounds) {
succ <- xgb.iter.update(bst$handle, dtrain, i - 1, obj) succ <- xgb.iter.update(bst$handle, dtrain, i - 1, obj)
if (length(watchlist) != 0) { if (length(watchlist) != 0) {
msg <- xgb.iter.eval(bst$handle, watchlist, i - 1, feval) msg <- xgb.iter.eval(bst$handle, watchlist, i - 1, feval)
if (0== ( (i-1) %% printEveryN)) if (0== ( (i-1) %% print.every.n))
cat(paste(msg, "\n", sep="")) cat(paste(msg, "\n", sep=""))
if (!is.null(early_stop_round)) if (!is.null(early.stop.round))
{ {
score = strsplit(msg,':|\\s+')[[1]][3] score = strsplit(msg,':|\\s+')[[1]][3]
score = as.numeric(score) score = as.numeric(score)
@ -205,7 +201,7 @@ xgb.train <- function(params=list(), data, nrounds, watchlist = list(),
bestScore = score bestScore = score
bestInd = i bestInd = i
} else { } else {
if (i-bestInd>=early_stop_round) { if (i-bestInd>=early.stop.round) {
earlyStopflag = TRUE earlyStopflag = TRUE
cat('Stopping. Best iteration:',bestInd) cat('Stopping. Best iteration:',bestInd)
break break
@ -215,7 +211,7 @@ xgb.train <- function(params=list(), data, nrounds, watchlist = list(),
} }
} }
bst <- xgb.Booster.check(bst) bst <- xgb.Booster.check(bst)
if (!is.null(early_stop_round)) { if (!is.null(early.stop.round)) {
bst$bestScore = bestScore bst$bestScore = bestScore
bst$bestInd = bestInd bst$bestInd = bestInd
} }

12
R-package/R/xgboost.R
View File

@ -28,14 +28,13 @@
#' @param verbose If 0, xgboost will stay silent. If 1, xgboost will print #' @param verbose If 0, xgboost will stay silent. If 1, xgboost will print
#' information of performance. If 2, xgboost will print information of both #' information of performance. If 2, xgboost will print information of both
#' performance and construction progress information #' performance and construction progress information
#' @param printEveryN Print every N progress messages when \code{verbose>0}. Default is 1 which means all messages are printed. #' @param print.every.n Print every N progress messages when \code{verbose>0}. Default is 1 which means all messages are printed.
#' @param missing Missing is only used when input is dense matrix, pick a float #' @param missing Missing is only used when input is dense matrix, pick a float
#' value that represents missing value. Sometimes a data use 0 or other extreme value to represents missing values. #' value that represents missing value. Sometimes a data use 0 or other extreme value to represents missing values.
#' @param early_stop_round If \code{NULL}, the early stopping function is not triggered. #' @param early.stop.round If \code{NULL}, the early stopping function is not triggered.
#' If set to an integer \code{k}, training with a validation set will stop if the performance #' If set to an integer \code{k}, training with a validation set will stop if the performance
#' keeps getting worse consecutively for \code{k} rounds. #' keeps getting worse consecutively for \code{k} rounds.
#' @param early.stop.round An alternative of \code{early_stop_round}. #' @param maximize If \code{feval} and \code{early.stop.round} are set, then \code{maximize} must be set as well.
#' @param maximize If \code{feval} and \code{early_stop_round} are set, then \code{maximize} must be set as well.
#' \code{maximize=TRUE} means the larger the evaluation score the better. #' \code{maximize=TRUE} means the larger the evaluation score the better.
#' @param ... other parameters to pass to \code{params}. #' @param ... other parameters to pass to \code{params}.
#' #'
@ -58,7 +57,7 @@
#' @export #' @export
#' #'
xgboost <- function(data = NULL, label = NULL, missing = NULL, params = list(), nrounds, xgboost <- function(data = NULL, label = NULL, missing = NULL, params = list(), nrounds,
verbose = 1, printEveryN=1L, early_stop_round = NULL, early.stop.round = NULL, verbose = 1, print.every.n = 1L, early.stop.round = NULL,
maximize = NULL, ...) { maximize = NULL, ...) {
if (is.null(missing)) { if (is.null(missing)) {
dtrain <- xgb.get.DMatrix(data, label) dtrain <- xgb.get.DMatrix(data, label)
@ -74,8 +73,7 @@ xgboost <- function(data = NULL, label = NULL, missing = NULL, params = list(),
watchlist <- list() watchlist <- list()
} }
bst <- xgb.train(params, dtrain, nrounds, watchlist, verbose = verbose, printEveryN=printEveryN, bst <- xgb.train(params, dtrain, nrounds, watchlist, verbose = verbose, print.every.n=print.every.n,
early_stop_round = early_stop_round,
early.stop.round = early.stop.round) early.stop.round = early.stop.round)
return(bst) return(bst)

12
R-package/man/xgb.cv.Rd
View File

@ -7,8 +7,8 @@
xgb.cv(params = list(), data, nrounds, nfold, label = NULL, xgb.cv(params = list(), data, nrounds, nfold, label = NULL,
missing = NULL, prediction = FALSE, showsd = TRUE, metrics = list(), missing = NULL, prediction = FALSE, showsd = TRUE, metrics = list(),
obj = NULL, feval = NULL, stratified = TRUE, folds = NULL, obj = NULL, feval = NULL, stratified = TRUE, folds = NULL,
verbose = T, printEveryN = 1L, early_stop_round = NULL, verbose = T, print.every.n = 1L, early.stop.round = NULL,
early.stop.round = NULL, maximize = NULL, ...) maximize = NULL, ...)
} }
\arguments{ \arguments{
\item{params}{the list of parameters. Commonly used ones are: \item{params}{the list of parameters. Commonly used ones are:
@ -66,15 +66,13 @@ If folds are supplied, the nfold and stratified parameters would be ignored.}
\item{verbose}{\code{boolean}, print the statistics during the process} \item{verbose}{\code{boolean}, print the statistics during the process}
\item{printEveryN}{Print every N progress messages when \code{verbose>0}. Default is 1 which means all messages are printed.} \item{print.every.n}{Print every N progress messages when \code{verbose>0}. Default is 1 which means all messages are printed.}
\item{early_stop_round}{If \code{NULL}, the early stopping function is not triggered. \item{early.stop.round}{If \code{NULL}, the early stopping function is not triggered.
If set to an integer \code{k}, training with a validation set will stop if the performance If set to an integer \code{k}, training with a validation set will stop if the performance
keeps getting worse consecutively for \code{k} rounds.} keeps getting worse consecutively for \code{k} rounds.}
\item{early.stop.round}{An alternative of \code{early_stop_round}.} \item{maximize}{If \code{feval} and \code{early.stop.round} are set, then \code{maximize} must be set as well.
\item{maximize}{If \code{feval} and \code{early_stop_round} are set, then \code{maximize} must be set as well.
\code{maximize=TRUE} means the larger the evaluation score the better.} \code{maximize=TRUE} means the larger the evaluation score the better.}
\item{...}{other parameters to pass to \code{params}.} \item{...}{other parameters to pass to \code{params}.}

10
R-package/man/xgb.train.Rd
View File

@ -5,7 +5,7 @@
\title{eXtreme Gradient Boosting Training} \title{eXtreme Gradient Boosting Training}
\usage{ \usage{
xgb.train(params = list(), data, nrounds, watchlist = list(), obj = NULL, xgb.train(params = list(), data, nrounds, watchlist = list(), obj = NULL,
feval = NULL, verbose = 1, printEveryN = 1L, early_stop_round = NULL, feval = NULL, verbose = 1, print.every.n = 1L,
early.stop.round = NULL, maximize = NULL, ...) early.stop.round = NULL, maximize = NULL, ...)
} }
\arguments{ \arguments{
@ -78,15 +78,13 @@ prediction and dtrain,}
\item{verbose}{If 0, xgboost will stay silent. If 1, xgboost will print \item{verbose}{If 0, xgboost will stay silent. If 1, xgboost will print
information of performance. If 2, xgboost will print information of both} information of performance. If 2, xgboost will print information of both}
\item{printEveryN}{Print every N progress messages when \code{verbose>0}. Default is 1 which means all messages are printed.} \item{print.every.n}{Print every N progress messages when \code{verbose>0}. Default is 1 which means all messages are printed.}
\item{early_stop_round}{If \code{NULL}, the early stopping function is not triggered. \item{early.stop.round}{If \code{NULL}, the early stopping function is not triggered.
If set to an integer \code{k}, training with a validation set will stop if the performance If set to an integer \code{k}, training with a validation set will stop if the performance
keeps getting worse consecutively for \code{k} rounds.} keeps getting worse consecutively for \code{k} rounds.}
\item{early.stop.round}{An alternative of \code{early_stop_round}.} \item{maximize}{If \code{feval} and \code{early.stop.round} are set, then \code{maximize} must be set as well.
\item{maximize}{If \code{feval} and \code{early_stop_round} are set, then \code{maximize} must be set as well.
\code{maximize=TRUE} means the larger the evaluation score the better.} \code{maximize=TRUE} means the larger the evaluation score the better.}
\item{...}{other parameters to pass to \code{params}.} \item{...}{other parameters to pass to \code{params}.}

12
R-package/man/xgboost.Rd
View File

@ -5,8 +5,8 @@
\title{eXtreme Gradient Boosting (Tree) library} \title{eXtreme Gradient Boosting (Tree) library}
\usage{ \usage{
xgboost(data = NULL, label = NULL, missing = NULL, params = list(), xgboost(data = NULL, label = NULL, missing = NULL, params = list(),
nrounds, verbose = 1, printEveryN = 1L, early_stop_round = NULL, nrounds, verbose = 1, print.every.n = 1L, early.stop.round = NULL,
early.stop.round = NULL, maximize = NULL, ...) maximize = NULL, ...)
} }
\arguments{ \arguments{
\item{data}{takes \code{matrix}, \code{dgCMatrix}, local data file or \item{data}{takes \code{matrix}, \code{dgCMatrix}, local data file or
@ -42,15 +42,13 @@ Commonly used ones are:
information of performance. If 2, xgboost will print information of both information of performance. If 2, xgboost will print information of both
performance and construction progress information} performance and construction progress information}
\item{printEveryN}{Print every N progress messages when \code{verbose>0}. Default is 1 which means all messages are printed.} \item{print.every.n}{Print every N progress messages when \code{verbose>0}. Default is 1 which means all messages are printed.}
\item{early_stop_round}{If \code{NULL}, the early stopping function is not triggered. \item{early.stop.round}{If \code{NULL}, the early stopping function is not triggered.
If set to an integer \code{k}, training with a validation set will stop if the performance If set to an integer \code{k}, training with a validation set will stop if the performance
keeps getting worse consecutively for \code{k} rounds.} keeps getting worse consecutively for \code{k} rounds.}
\item{early.stop.round}{An alternative of \code{early_stop_round}.} \item{maximize}{If \code{feval} and \code{early.stop.round} are set, then \code{maximize} must be set as well.
\item{maximize}{If \code{feval} and \code{early_stop_round} are set, then \code{maximize} must be set as well.
\code{maximize=TRUE} means the larger the evaluation score the better.} \code{maximize=TRUE} means the larger the evaluation score the better.}
\item{...}{other parameters to pass to \code{params}.} \item{...}{other parameters to pass to \code{params}.}