Extreme Gradient Boosting Trainer

Trains a XGBoost model in R

Details

Trains a Extreme Gradient Boosting Model. XGBoost belongs to a family of boosting algorithms that creates an ensemble of weak learner to learn about data. It is a wrapper for original xgboost R package, you can find the documentation here: http://xgboost.readthedocs.io/en/latest/parameter.html

Public fields

booster: the trainer type, the values are gbtree(default), gblinear, dart:gbtree
objective: specify the learning task. Check the link above for all possible values.
nthread: number of parallel threads used to run, default is to run using all threads available
silent: 0 means printing running messages, 1 means silent mode
n_estimators: number of trees to grow, default = 100
learning_rate: Step size shrinkage used in update to prevents overfitting. Lower the learning rate, more time it takes in training, value lies between between 0 and 1. Default = 0.3
gamma: Minimum loss reduction required to make a further partition on a leaf node of the tree. The larger gamma is, the more conservative the algorithm will be. Value lies between 0 and infinity, Default = 0
max_depth: the maximum depth of each tree, default = 6
min_child_weight: Minimum sum of instance weight (hessian) needed in a child. If the tree partition step results in a leaf node with the sum of instance weight less than min_child_weight, then the building process will give up further partitioning. In linear regression task, this simply corresponds to minimum number of instances needed to be in each node. The larger min_child_weight is, the more conservative the algorithm will be. Value lies between 0 and infinity. Default = 1
subsample: Subsample ratio of the training instances. Setting it to 0.5 means that XGBoost would randomly sample half of the training data prior to growing trees. and this will prevent overfitting. Subsampling will occur once in every boosting iteration. Value lies between 0 and 1. Default = 1
colsample_bytree: Subsample ratio of columns when constructing each tree. Subsampling will occur once in every boosting iteration. Value lies between 0 and 1. Default = 1
lambda: L2 regularization term on weights. Increasing this value will make model more conservative. Default = 1
alpha: L1 regularization term on weights. Increasing this value will make model more conservative. Default = 0
eval_metric: Evaluation metrics for validation data, a default metric will be assigned according to objective
print_every: print training log after n iterations. Default = 50
feval: custom evaluation function
early_stopping: Used to prevent overfitting, stops model training after this number of iterations if there is no improvement seen
maximize: If feval and early_stopping_rounds are set, then this parameter must be set as well. When it is TRUE, it means the larger the evaluation score the better.
custom_objective: custom objective function
save_period: when it is non-NULL, model is saved to disk after every save_period rounds, 0 means save at the end.
save_name: the name or path for periodically saved model file.
xgb_model: a previously built model to continue the training from. Could be either an object of class xgb.Booster, or its raw data, or the name of a file with a previously saved model.
callbacks: a list of callback functions to perform various task during boosting. See callbacks. Some of the callbacks are automatically created depending on the parameters' values. User can provide either existing or their own callback methods in order to customize the training process.
verbose: If 0, xgboost will stay silent. If 1, xgboost will print information of performance. If 2, xgboost will print some additional information. Setting verbose > 0 automatically engages the cb.evaluation.log and cb.print.evaluation callback functions.
watchlist: what information should be printed when verbose=1 or verbose=2. Watchlist is used to specify validation set monitoring during training. For example user can specify watchlist=list(validation1=mat1, validation2=mat2) to watch the performance of each round's model on mat1 and mat2
num_class: set number of classes in case of multiclassification problem
weight: a vector indicating the weight for each row of the input.
na_missing: by default is set to NA, which means that NA values should be considered as 'missing' by the algorithm. Sometimes, 0 or other extreme value might be used to represent missing values. This parameter is only used when input is a dense matrix.
feature_names: internal use, stores the feature names for model importance
cv_model: internal use

Methods

Method `new()`

Usage

XGBTrainer$new(
  booster,
  objective,
  nthread,
  silent,
  n_estimators,
  learning_rate,
  gamma,
  max_depth,
  min_child_weight,
  subsample,
  colsample_bytree,
  lambda,
  alpha,
  eval_metric,
  print_every,
  feval,
  early_stopping,
  maximize,
  custom_objective,
  save_period,
  save_name,
  xgb_model,
  callbacks,
  verbose,
  num_class,
  weight,
  na_missing
)

Arguments

booster: the trainer type, the values are gbtree(default), gblinear, dart:gbtree
objective: specify the learning task. Check the link above for all possible values.
nthread: number of parallel threads used to run, default is to run using all threads available
silent: 0 means printing running messages, 1 means silent mode
n_estimators: number of trees to grow, default = 100
learning_rate: Step size shrinkage used in update to prevents overfitting. Lower the learning rate, more time it takes in training, value lies between between 0 and 1. Default = 0.3
gamma: Minimum loss reduction required to make a further partition on a leaf node of the tree. The larger gamma is, the more conservative the algorithm will be. Value lies between 0 and infinity, Default = 0
max_depth: the maximum depth of each tree, default = 6
min_child_weight: Minimum sum of instance weight (hessian) needed in a child. If the tree partition step results in a leaf node with the sum of instance weight less than min_child_weight, then the building process will give up further partitioning. In linear regression task, this simply corresponds to minimum number of instances needed to be in each node. The larger min_child_weight is, the more conservative the algorithm will be. Value lies between 0 and infinity. Default = 1
subsample: Subsample ratio of the training instances. Setting it to 0.5 means that XGBoost would randomly sample half of the training data prior to growing trees. and this will prevent overfitting. Subsampling will occur once in every boosting iteration. Value lies between 0 and 1. Default = 1
colsample_bytree: Subsample ratio of columns when constructing each tree. Subsampling will occur once in every boosting iteration. Value lies between 0 and 1. Default = 1
lambda: L2 regularization term on weights. Increasing this value will make model more conservative. Default = 1
alpha: L1 regularization term on weights. Increasing this value will make model more conservative. Default = 0
eval_metric: Evaluation metrics for validation data, a default metric will be assigned according to objective
print_every: print training log after n iterations. Default = 50
feval: custom evaluation function
early_stopping: Used to prevent overfitting, stops model training after this number of iterations if there is no improvement seen
maximize: If feval and early_stopping_rounds are set, then this parameter must be set as well. When it is TRUE, it means the larger the evaluation score the better.
custom_objective: custom objective function
save_period: when it is non-NULL, model is saved to disk after every save_period rounds, 0 means save at the end.
save_name: the name or path for periodically saved model file.
xgb_model: a previously built model to continue the training from. Could be either an object of class xgb.Booster, or its raw data, or the name of a file with a previously saved model.
callbacks: a list of callback functions to perform various task during boosting. See callbacks. Some of the callbacks are automatically created depending on the parameters' values. User can provide either existing or their own callback methods in order to customize the training process.
verbose: If 0, xgboost will stay silent. If 1, xgboost will print information of performance. If 2, xgboost will print some additional information. Setting verbose > 0 automatically engages the cb.evaluation.log and cb.print.evaluation callback functions.
num_class: set number of classes in case of multiclassification problem
weight: a vector indicating the weight for each row of the input.
na_missing: by default is set to NA, which means that NA values should be considered as 'missing' by the algorithm. Sometimes, 0 or other extreme value might be used to represent missing values. This parameter is only used when input is a dense matrix.

Details

Create a new `XGBTrainer` object.

Returns

A `XGBTrainer` object.

Examples

library(data.table)
df <- copy(iris)

# convert characters/factors to numeric
df$Species <- as.numeric(as.factor(df$Species))-1

# initialise model
xgb <- XGBTrainer$new(objective = 'multi:softmax',
                      maximize = FALSE,
                      eval_metric = 'merror',
                      num_class=3,
                      n_estimators = 2)

Method `cross_val()`

Usage

XGBTrainer$cross_val(X, y, nfolds = 5, stratified = TRUE, folds = NULL)

Arguments

X: data.frame
y: character, name of target variable
nfolds: integer, number of folds
stratified: logical, whether to use stratified sampling
folds: the list of CV folds' indices - either those passed through the folds parameter or randomly generated.

Details

Trains the xgboost model using cross validation scheme

Returns

NULL, trains a model and saves it in memory

Examples

\dontrun{
library(data.table)
df <- copy(iris)

# convert characters/factors to numeric
df$Species <- as.numeric(as.factor(df$Species))-1

# initialise model
xgb <- XGBTrainer$new(objective = 'multi:softmax',
                      maximize = FALSE,
                      eval_metric = 'merror',
                      num_class=3,
                      n_estimators = 2)

# do cross validation to find optimal value for n_estimators
xgb$cross_val(X = df, y = 'Species',nfolds = 3, stratified = TRUE)
}

Method `fit()`

Usage

XGBTrainer$fit(X, y, valid = NULL)

Arguments

X: data.frame, training data
y: character, name of target variable
valid: data.frame, validation data

Details

Fits the xgboost model on given data

Returns

NULL, trains a model and keeps it in memory

Examples

library(data.table)
df <- copy(iris)

# convert characters/factors to numeric
df$Species <- as.numeric(as.factor(df$Species))-1

# initialise model
xgb <- XGBTrainer$new(objective = 'multi:softmax',
                      maximize = FALSE,
                      eval_metric = 'merror',
                      num_class=3,
                      n_estimators = 2)
xgb$fit(df, 'Species')

Method `predict()`

Usage

XGBTrainer$predict(df)

Arguments

df: data.frame, test data set

Details

Returns predicted values for a given test data

Returns

xgboost predictions

Examples

#' library(data.table)
df <- copy(iris)

# convert characters/factors to numeric
df$Species <- as.numeric(as.factor(df$Species))-1

# initialise model
xgb <- XGBTrainer$new(objective = 'multi:softmax',
                      maximize = FALSE,
                      eval_metric = 'merror',
                      num_class=3,
                      n_estimators = 2)
xgb$fit(df, 'Species')

# make predictions
preds <- xgb$predict(as.matrix(iris[,1:4]))

Method `show_importance()`

Usage

XGBTrainer$show_importance(type = "plot", topn = 10)

Arguments

type: character, could be 'plot' or 'table'
topn: integer, top n features to display

Details

Shows feature importance plot

Returns

a table or a plot of feature importance

Examples

\dontrun{
library(data.table)
df <- copy(iris)

# convert characters/factors to numeric
df$Species <- as.numeric(as.factor(df$Species))-1

# initialise model
xgb <- XGBTrainer$new(objective = 'multi:softmax',
                      maximize = FALSE,
                      eval_metric = 'merror',
                      num_class=3,
                      n_estimators = 2)
xgb$fit(df, 'Species')
xgb$show_importance()
}

Method `clone()`

The objects of this class are cloneable with this method.

Usage

XGBTrainer$clone(deep = FALSE)

Arguments

deep: Whether to make a deep clone.

Examples


## ------------------------------------------------
## Method `XGBTrainer$new`
## ------------------------------------------------

library(data.table)
df <- copy(iris)

# convert characters/factors to numeric
df$Species <- as.numeric(as.factor(df$Species))-1

# initialise model
xgb <- XGBTrainer$new(objective = 'multi:softmax',
                      maximize = FALSE,
                      eval_metric = 'merror',
                      num_class=3,
                      n_estimators = 2)

## ------------------------------------------------
## Method `XGBTrainer$cross_val`
## ------------------------------------------------

if (FALSE) {
library(data.table)
df <- copy(iris)

# convert characters/factors to numeric
df$Species <- as.numeric(as.factor(df$Species))-1

# initialise model
xgb <- XGBTrainer$new(objective = 'multi:softmax',
                      maximize = FALSE,
                      eval_metric = 'merror',
                      num_class=3,
                      n_estimators = 2)

# do cross validation to find optimal value for n_estimators
xgb$cross_val(X = df, y = 'Species',nfolds = 3, stratified = TRUE)
}

## ------------------------------------------------
## Method `XGBTrainer$fit`
## ------------------------------------------------

library(data.table)
df <- copy(iris)

# convert characters/factors to numeric
df$Species <- as.numeric(as.factor(df$Species))-1

# initialise model
xgb <- XGBTrainer$new(objective = 'multi:softmax',
                      maximize = FALSE,
                      eval_metric = 'merror',
                      num_class=3,
                      n_estimators = 2)
xgb$fit(df, 'Species')
#> converting the data into xgboost format..
#> starting with training...
#> [10:48:34] WARNING: amalgamation/../src/learner.cc:627: 
#> Parameters: { "nrounds" } might not be used.
#> 
#>   This could be a false alarm, with some parameters getting used by language bindings but
#>   then being mistakenly passed down to XGBoost core, or some parameter actually being used
#>   but getting flagged wrongly here. Please open an issue if you find any such cases.
#> 
#> 
#> [1]	train-merror:0.020000 
#> Will train until train_merror hasn't improved in 50 rounds.
#> 
#> [2]	train-merror:0.026667 

## ------------------------------------------------
## Method `XGBTrainer$predict`
## ------------------------------------------------

#' library(data.table)
df <- copy(iris)

# convert characters/factors to numeric
df$Species <- as.numeric(as.factor(df$Species))-1

# initialise model
xgb <- XGBTrainer$new(objective = 'multi:softmax',
                      maximize = FALSE,
                      eval_metric = 'merror',
                      num_class=3,
                      n_estimators = 2)
xgb$fit(df, 'Species')
#> converting the data into xgboost format..
#> starting with training...
#> [10:48:34] WARNING: amalgamation/../src/learner.cc:627: 
#> Parameters: { "nrounds" } might not be used.
#> 
#>   This could be a false alarm, with some parameters getting used by language bindings but
#>   then being mistakenly passed down to XGBoost core, or some parameter actually being used
#>   but getting flagged wrongly here. Please open an issue if you find any such cases.
#> 
#> 
#> [1]	train-merror:0.020000 
#> Will train until train_merror hasn't improved in 50 rounds.
#> 
#> [2]	train-merror:0.026667 

# make predictions
preds <- xgb$predict(as.matrix(iris[,1:4]))

## ------------------------------------------------
## Method `XGBTrainer$show_importance`
## ------------------------------------------------

if (FALSE) {
library(data.table)
df <- copy(iris)

# convert characters/factors to numeric
df$Species <- as.numeric(as.factor(df$Species))-1

# initialise model
xgb <- XGBTrainer$new(objective = 'multi:softmax',
                      maximize = FALSE,
                      eval_metric = 'merror',
                      num_class=3,
                      n_estimators = 2)
xgb$fit(df, 'Species')
xgb$show_importance()
}

Details

Public fields

Methods

Public methods

Method new()

Usage

Arguments

Details

Returns

Examples

Method cross_val()

Usage

Arguments

Details

Returns

Examples

Method fit()

Usage

Arguments

Details

Returns

Examples

Method predict()

Usage

Arguments

Details

Returns

Examples

Method show_importance()

Usage

Arguments

Details

Returns

Examples

Method clone()

Usage

Arguments

Examples

Method `new()`

Method `cross_val()`

Method `fit()`

Method `predict()`

Method `show_importance()`

Method `clone()`