This class is the raw C++ pendant and still at a very high-level.
It is the base for the Compboost R6 class and provides
many convenient wrapper to access data and execute methods by calling
the C++ methods.
Format
S4 object.
Arguments
- oob_response
(ResponseRegr | ResponseBinaryClassif)
The response object containing the target variable.- learning_rate
(
numeric(1))
The learning rate.- stop_if_all_stopper_fulfilled
(
logical(1))
Boolean to indicate which stopping strategy is used. IfTRUE, the algorithm stops if the conditions of all loggers for stopping apply.- factory_list
(BlearnerFactoryList)
List of base learner factories from which one base learner is selected in each iteration by using the- loss
(LossQuadratic | LossBinomial | LossHuber | LossAbsolute | LossQuantile)
An initializedS4loss object (requires to callLoss*$new(...)). See the respective help page for further information.- logger_list
(LoggerList)
The LoggerList object with all loggers.- optimizer
(OptimizerCoordinateDescent | OptimizerCoordinateDescentLineSearch | OptimizerAGBM | OptimizerCosineAnnealing)
An initializedS4optimizer object (requires to callOptimizer*.new(..). See the respective help page for further information.
Usage
Compboost$new(response, learning_rate, stop_if_all_stopper_fulfilled,
factory_list, loss, logger_list, optimizer)Methods
$train():() -> ()$continueTraining():() -> ()$getLearningRate():() -> numeric(1)$getPrediction():() -> matrix()$getSelectedBaselearner():() -> character()$getLoggerData():() -> list(character(), matrix())$getEstimatedParameter():() -> list(matrix())$getParameterAtIteration():() -> list(matrix())$getParameterMatrix():() -> matrix()$predictFactoryTrainData():() -> matrix()$predictFactoryNewData():list(Data*) -> matrix()$predictIndividualTrainData():() -> list(matrix())Get the linear contribution of each base learner for the training data.$predictIndividual():list(Data*) -> list(matrix())Get the linear contribution of each base learner for new data.$predict():list(Data*), logical(1) -> matrix()$summarizeCompboost():() -> ()$isTrained():() -> logical(1)$setToIteration():() -> ()$saveJson():() -> ()$getOffset():() -> numeric(1) | matrix()$getRiskVector():() -> numeric()$getResponse():() -> Response*$getOptimizer():() -> Optimizer*$getLoss():() -> Loss*$getLoggerList():() -> LoggerList$getBaselearnerList():() -> BlearnerFactoryList$useGlobalStopping():() -> logical(1)*$getFactoryMap():() -> list(Baselearner*)$getDataMap():() -> list(Data*)
Examples
# Some data:
df = mtcars
df$mpg_cat = ifelse(df$mpg > 20, "high", "low")
# # Create new variable to check the polynomial base learner with degree 2:
# df$hp2 = df[["hp"]]^2
# Data for the baselearner are matrices:
X_hp = as.matrix(df[["hp"]])
X_wt = as.matrix(df[["wt"]])
# Target variable:
response = ResponseBinaryClassif$new("mpg_cat", "high", df[["mpg_cat"]])
data_source_hp = InMemoryData$new(X_hp, "hp")
data_source_wt = InMemoryData$new(X_wt, "wt")
# List for oob logging:
oob_data = list(data_source_hp, data_source_wt)
# List to test prediction on newdata:
test_data = oob_data
# Factories:
linear_factory_hp = BaselearnerPolynomial$new(data_source_hp,
list(degree = 1, intercept = TRUE))
linear_factory_wt = BaselearnerPolynomial$new(data_source_wt,
list(degree = 1, intercept = TRUE))
quadratic_factory_hp = BaselearnerPolynomial$new(data_source_hp,
list(degree = 2, intercept = TRUE))
spline_factory_wt = BaselearnerPSpline$new(data_source_wt,
list(degree = 3, n_knots = 10, penalty = 2, differences = 2))
# Create new factory list:
factory_list = BlearnerFactoryList$new()
# Register factories:
factory_list$registerFactory(linear_factory_hp)
factory_list$registerFactory(linear_factory_wt)
factory_list$registerFactory(quadratic_factory_hp)
factory_list$registerFactory(spline_factory_wt)
# Define loss:
loss_bin = LossBinomial$new()
# Define optimizer:
optimizer = OptimizerCoordinateDescent$new()
## Logger
# Define logger. We want just the iterations as stopper but also track the
# time, inbag risk and oob risk:
log_iterations = LoggerIteration$new(" iteration_logger", TRUE, 500)
log_time = LoggerTime$new("time_logger", FALSE, 500, "microseconds")
# Define new logger list:
logger_list = LoggerList$new()
# Register the logger:
logger_list$registerLogger(log_iterations)
logger_list$registerLogger(log_time)
# Run compboost:
# --------------
# Initialize object:
cboost = Compboost_internal$new(
response = response,
learning_rate = 0.05,
stop_if_all_stopper_fulfilled = FALSE,
factory_list = factory_list,
loss = loss_bin,
logger_list = logger_list,
optimizer = optimizer
)
# Train the model (we want to print the trace):
cboost$train(trace = 50)
#> 1/500 risk = 0.68 time_logger = 0
#> 50/500 risk = 0.42 time_logger = 635
#> 100/500 risk = 0.31 time_logger = 1288
#> 150/500 risk = 0.25 time_logger = 2056
#> 200/500 risk = 0.21 time_logger = 2914
#> 250/500 risk = 0.19 time_logger = 3897
#> 300/500 risk = 0.17 time_logger = 4933
#> 350/500 risk = 0.16 time_logger = 6079
#> 400/500 risk = 0.15 time_logger = 7334
#> 450/500 risk = 0.14 time_logger = 8631
#> 500/500 risk = 0.13 time_logger = 10002
#>
#>
#> Train 500 iterations in 0 Seconds.
#> Final risk based on the train set: 0.13
#>
cboost
#>
#> Compboost object with:
#> - Learning Rate: 0.05
#> - Are all logger used as stopper: 0
#> - Model is already trained with 500 iterations/fitted baselearner
#> - Actual state is at iteration 500
#>
#>
#>
# Get estimated parameter:
cboost$getEstimatedParameter()
#> $hp_poly2
#> [,1]
#> [1,] 6.0660349854
#> [2,] -0.0645193461
#> [3,] 0.0001267875
#>
#> $wt_spline_degree_3
#> [,1]
#> [1,] 1.9370515
#> [2,] 1.8485620
#> [3,] 1.8081122
#> [4,] 1.8647942
#> [5,] 1.7149991
#> [6,] 1.0827258
#> [7,] -0.4203328
#> [8,] -1.8130157
#> [9,] -2.0641036
#> [10,] -2.0050272
#> [11,] -1.8182042
#> [12,] -1.5569128
#> [13,] -1.2730290
#> [14,] -0.9904981
#>
# Get trace of selected base learner:
cboost$getSelectedBaselearner()
#> [1] "wt_spline_degree_3" "wt_spline_degree_3" "wt_spline_degree_3"
#> [4] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [7] "hp_poly2" "wt_spline_degree_3" "wt_spline_degree_3"
#> [10] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [13] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [16] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [19] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [22] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [25] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [28] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [31] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [34] "wt_spline_degree_3" "wt_spline_degree_3" "hp_poly2"
#> [37] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [40] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [43] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [46] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [49] "wt_spline_degree_3" "wt_spline_degree_3" "hp_poly2"
#> [52] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [55] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [58] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [61] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [64] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [67] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [70] "wt_spline_degree_3" "wt_spline_degree_3" "hp_poly2"
#> [73] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [76] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [79] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [82] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [85] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [88] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [91] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [94] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [97] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [100] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [103] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [106] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [109] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [112] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [115] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [118] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [121] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [124] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [127] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [130] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [133] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [136] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [139] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [142] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [145] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [148] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [151] "wt_spline_degree_3" "hp_poly2" "hp_poly2"
#> [154] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [157] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [160] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [163] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [166] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [169] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [172] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [175] "hp_poly2" "hp_poly2" "wt_spline_degree_3"
#> [178] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [181] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [184] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [187] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [190] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [193] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [196] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [199] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [202] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [205] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [208] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [211] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [214] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [217] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [220] "hp_poly2" "hp_poly2" "wt_spline_degree_3"
#> [223] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [226] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [229] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [232] "wt_spline_degree_3" "hp_poly2" "hp_poly2"
#> [235] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [238] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [241] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [244] "hp_poly2" "hp_poly2" "wt_spline_degree_3"
#> [247] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [250] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [253] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [256] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [259] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [262] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [265] "wt_spline_degree_3" "hp_poly2" "hp_poly2"
#> [268] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [271] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [274] "wt_spline_degree_3" "hp_poly2" "hp_poly2"
#> [277] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [280] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [283] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [286] "hp_poly2" "hp_poly2" "wt_spline_degree_3"
#> [289] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [292] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [295] "hp_poly2" "hp_poly2" "wt_spline_degree_3"
#> [298] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [301] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [304] "hp_poly2" "hp_poly2" "wt_spline_degree_3"
#> [307] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [310] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [313] "hp_poly2" "hp_poly2" "wt_spline_degree_3"
#> [316] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [319] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [322] "hp_poly2" "hp_poly2" "wt_spline_degree_3"
#> [325] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [328] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [331] "hp_poly2" "hp_poly2" "wt_spline_degree_3"
#> [334] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [337] "wt_spline_degree_3" "hp_poly2" "hp_poly2"
#> [340] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [343] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [346] "wt_spline_degree_3" "hp_poly2" "hp_poly2"
#> [349] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [352] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [355] "wt_spline_degree_3" "hp_poly2" "hp_poly2"
#> [358] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [361] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [364] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [367] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [370] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [373] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [376] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [379] "hp_poly2" "hp_poly2" "wt_spline_degree_3"
#> [382] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [385] "wt_spline_degree_3" "hp_poly2" "hp_poly2"
#> [388] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [391] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [394] "wt_spline_degree_3" "hp_poly2" "hp_poly2"
#> [397] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [400] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [403] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [406] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [409] "hp_poly2" "hp_poly2" "wt_spline_degree_3"
#> [412] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [415] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [418] "hp_poly2" "hp_poly2" "wt_spline_degree_3"
#> [421] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [424] "wt_spline_degree_3" "hp_poly2" "hp_poly2"
#> [427] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [430] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [433] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [436] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [439] "hp_poly2" "hp_poly2" "wt_spline_degree_3"
#> [442] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [445] "wt_spline_degree_3" "hp_poly2" "hp_poly2"
#> [448] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [451] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [454] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [457] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [460] "hp_poly2" "hp_poly2" "wt_spline_degree_3"
#> [463] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [466] "wt_spline_degree_3" "hp_poly2" "hp_poly2"
#> [469] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [472] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [475] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [478] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [481] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [484] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [487] "wt_spline_degree_3" "hp_poly2" "wt_spline_degree_3"
#> [490] "hp_poly2" "hp_poly2" "wt_spline_degree_3"
#> [493] "hp_poly2" "wt_spline_degree_3" "hp_poly2"
#> [496] "wt_spline_degree_3" "hp_poly2" "hp_poly2"
#> [499] "wt_spline_degree_3" "hp_poly2"
# Set to iteration 200:
cboost$setToIteration(200, 30)
# Get new parameter values:
cboost$getEstimatedParameter()
#> $hp_poly2
#> [,1]
#> [1,] 3.649127e+00
#> [2,] -3.820112e-02
#> [3,] 7.376905e-05
#>
#> $wt_spline_degree_3
#> [,1]
#> [1,] 1.3479210
#> [2,] 1.3418629
#> [3,] 1.3528398
#> [4,] 1.3714923
#> [5,] 1.2096167
#> [6,] 0.7153452
#> [7,] -0.2712226
#> [8,] -1.1840205
#> [9,] -1.4335470
#> [10,] -1.4475909
#> [11,] -1.3450002
#> [12,] -1.1727850
#> [13,] -0.9769323
#> [14,] -0.7809478
#>