BootControl
, OOBControl
, and SplitControl
.SplitControl
.MLModel
objects without a na.rm
slot.role_binom()
, role_case()
, and role_surv()
to remove the requirement that their variables be present in newdata
supplied to predict()
.na.rm
to MLModel()
for construction of a model that automatically removes all cases with missing values from model fitting and prediction, none, or only those whose missing values are in the response variable. Set the na.rm
values in supplied MLModels
to automatically remove cases with missing values if not supported by their model fitting and prediction functions.prob.model
to SVMModel()
.verbose
to fit()
and predict()
.Error in as.data.frame(x) : object 'x' not found
issue when fitting a BARTMachineModel
that started occurring with bartMachine
package version 1.2.7.ModeledInput
and rpp()
.na.rm
to MLModel
.method
to r2()
for calculation of Pearson or Spearman correlation.predict()
S4 method for MLModelFit
.MLModelFunction()
.as.MLInput()
methods for MLModelFit
and ModelSpecification
.as.MLModel()
method for ModelSpecification
.SelectedInput
terms.StackedModel
and SuperModel
..MachineShop
list attribute to MLModelFit
.mlmodel
in MLModelFit
to model
in .MachineShop
.input
in MLModel
to .MachineShop
..MachineShop
to the predict
and varimp
slot functions of MLModel
.TypeError
in dependence()
with numeric dummy variables from recipes.ModelRecipe
with retain = TRUE
for recipe steps that are skipped, for example, when test datasets are created.auc()
, pr_auc()
, and roc_auc()
for multiclass factor responses.select
to rfe()
.perf_stats
not found in optim()
.conf
to set_optim_bayes()
.StackedModel
and SuperModel
in ModelSpecification()
.SelectedModelSpecification
.ModeledInput
, ModeledFrame
, and ModeledRecipe
.TunedModeledRecipe
.fixed
from TunedModel()
.Grid()
.rpp()
to ppr()
.ModeledInput()
with ModelSpecification()
.NNetModel
model-specific variable importance.SurvRegModelFit
summary()
errorCVControl
when stratification or grouping size leads to construction of fewer than requested folds for cross-validation resampling.type
with options "glance"
and "tidy"
to summary.MLModelFit()
.print.Resample()
.ModelSpecification
.set_monitor()
: monitoring of resampling and optimizationset_optim_bayes()
: Bayesian optimization with a Gaussian process modelset_optim_bfgs()
: low-memory quasi-Newton BFGS optimizationset_optim_grid()
: exhaustive and random grid searchesset_optim_method()
: user-defined optimization functionsset_optim_pso()
: particle swarm optimizationset_optim_sann()
: simulated annealingperformance()
method for MLModel
to replicate the previous behavior of summary.MLModel()
.performance()
, plot()
, and summary()
methods for TrainingStep
.Resample
performances.type
of predict()
."default"
for model-specific default predictions."numeric"
for numeric predictions."prob"
to be for probabilities between 0 and 1.confusion()
default behavior to convert factor probabilities to levels.control
to object
in set functions.f
to fun
in roc_index()
.ListOf
training step summaries from summary.MLModel()
.TrainingStep
object from rfe()
.expand_params()
.EnsembleModel
.MLOptimization
, GridSearch
, NullOptimization
, RandomGridSearch
, and SequentialOptimization
.NullControl
.control
to PerformanceCurve
.method
to TrainingStep
.optim
to TrainingParams
.params
to MLInput
.SelectedModel
from EnsembleModel
.StackedModel
from EnsembleModel
.SuperModel
from StackedModel
.case_comps
to vars
in Resample
.grid
to log
in TrainingStep
.GLMModel
print.TrainingStep()
TunedModel()
terms.formula()
.distr
and method
to dependence()
.ParsnipModel()
for model specifications (model_spec
) from the parsnip package.rfe()
for recursive feature elimination.as.MLModel()
for model_spec
and ModeledInput
.as.MLModel()
method.metric
of auc()
.method
default from "model"
to "permute"
in varimp()
.ModelFrame
to an S4 class; generally requires explicit conversion to a data frame with as.data.frame()
in MLModel
fit
and predict
functions.stat.Trained
to stat.TrainingParams
.stats.VarImp
.ParsnipModel
.SurvTimes
.TrainingParams
.Grid
.Params
.name
, selected
, and metrics
to slot grid
of TrainingStep
class.grid
to TunedInput
.id
to MLInput
and MLModel
classes.id
and name
to TrainingStep
class.models
to SelectedModel
.name
from MLControl
classes.selected
, values
, and metric
from TrainingStep
class.shift
from VariableImportance
class.Grid
to TuningGrid
.Resamples
to Resample
.TrainStep
to TrainingStep
.VarImp
to VariableImportance
.MLControl
.MLBootControl
→ BootControl
MLBootOptimismControl
→ BootOptimismControl
MLCVControl
→ CVControl
MLCVOptimismControl
→ CVOptimismControl
MLOOBControl
→ OOBControl
MLSplitControl
→ SplitControl
MLTrainControl
→ TrainControl
Input
and Model
to params
in slot grid
of TrainingStep
class.Resample
to Iteration
in Resample
classx
to input
in MLModel
class.XGBModel
nrounds
from 1 to 100.nrounds
and max_depth
in automated grids for XGBDARTModel
and XGBTreeModel
.nrounds
, lambda
, and alpha
in automated grid for XGBLinearModel
.survival:aft
prediction.survival:cox
to survival:aft
.TrainingStep
objects and output.varimp()
.model
→ object
in TunedModel()
x
→ object
in expand_model()
x
→ formula
/input
/model
in expand_modelgrid()
, fit()
, ModelFrame()
, resample()
, rfe()
methodsx
→ formula
/object
/model
in ModeledInput()
methodsx
→ object
in ParameterGrid()
methodsx
→ control
in set_monitor()
, set_predict()
, set_strata()
x
→ object
in TunedInput()
Grid()
to TuningGrid()
.ModelFrame()
.MLModel
params
slots.na.rm
to dependence()
.stats.VarImp
for summary statistics to compute on permutation-based variable importance.varimp()
.t.test.PerformanceDiff()
.metric
to type
in varimp()
functions for BartMachineModel
, C50Model
, EarthModel
, RFSRCModel
, and XGBModel
.type
default to "nsubsets"
in EarthModel
varimp()
.cross_entropy()
numeric
method.f
in roc_index()
Surv
method.weights
to MLModel
classes.LMModel
for all response types.breaks
in calibration()
.max = Inf
arguments to print.default()
.ModelFrame()
arguments strata
and weights
in data
environment.Weight
of case weights to Resamples
data frame.values
column to get_values
in MLModel
gridinfo
slot.resample_progress
and resample_verbose
to set_monitor()
arguments progress
and verbose
.MLControl()
arguments strata_breaks
, strata_nunique
, strata_prop
, and strata_size
to set_strata()
arguments breaks
, nunique
, prop
, and size
.MLControl()
arguments times
, distr
, and method
to set_predict()
.%>%
operator.Resamples
objects.regular
to default
in MLModel
gridinfo slot.size
and random
arguments of ParameterGrid()
to match those of Grid()
.coeflearn
values in their defined order instead of at random in AdaBoostModel
.kernels
values in their defined order instead of at random in KNNModel
.splitrule
methods in RangerModel
.splitrule
values in their defined order instead of at random in RangerModel
.max.print
to print_max
.progress.resample
to resample_progress
.stat.train
to stat.Trained
.dist.Surv
to distr.SurvMeans
.dist.SurvProbs
to distr.SurvProbs
.strata_breaks
, strata_nunique
, strata_prop
and strata_size
arguments to MLControl()
constructor.strata_breaks
if numeric quantile bins are below strata_prop
and strata_size
.strata_prop
and strata_size
iteratively.strata_prop
and strata_size
iteratively.length
arguments from Grid()
and ParameterGrid()
.gridinfo
functions in MLModel()
.brier()
metric."fleming-harrington"
as a choice for the method
argument of predict()
and for the method.EmpiricalSurv
global setting, because it is a special case of the existing (default) "efron"
choice and thus not needed."rayleigh"
choice for the distr.Surv
and distr.SurvProbs
global settings.dist
argument to distr
in calibration()
, MLControl()
, predict()
, and r2()
.distr
argument to SurvEvents()
and SurvProbs()
.SurvMeans
class.SurvMeans
object.calibration()
and r2()
."terms"
predictor_encoding to "model.frame"
in MLModel
class.performance()
response type-specific methods to metrics
supplied as a single MLMetric
function.get_grid()
with expand_modelgrid()
.GLMNetModel
.MLModel
.traininfo
slot to train_steps
in MLModel
classes.retain
argument in prep()
.fixed
argument default NULL
to list()
in TunedModel()
.length
argument to size
in Grid()
and ParameterGrid()
.ParameterGrid()
.grid
slot with gridinfo
in MLModel
classes.Grid()
.get_grid()
function to extract model-defined tuning grids.trainbits
slot to traininfo
in MLModel
classes.RPartModel
cp
grid points from cptable
according to smallest cross-validation error (mean plus one standard deviation).Performance
diff()
method.RFSRCModel
.unMLModelFit()
function to revert an MLModelFit
object to its original class.options
argument to step_lincomp()
and step_sbf()
.step_sbf()
function for variable selection by filtering.step_kmedoids
objects from step_sbf
, and refactor methods.tidy()
column medoids
to selected
.tidy()
column names
to name
.tidy()
non-selected variable names to NA
.step_lincomp()
function for linear components variable reduction.step_kmeans
objects from step_lincomp
, and refactor methods.tidy()
column names
to name
.step_spca
objects from step_lincomp
, and refactor methods.tidy()
column value
to weight
.tidy()
column component
to name
.GBMModel
distribution to bernoulli, instead of multinomial, for binary responses.RHS.formula
for listing of operators and functions allowed on right-hand side of traditional formulas.step_kmedoids()
.XGBModel
, XGBDARTModel
, XGBLinearModel
, and XGBTreeModel
.NNetModel
linout
argument automatically according to the response variable type (numeric: TRUE
, other: FALSE
). Previously, linout
had a default value of FALSE
as defined in the nnet
package.NNetModel
fit()
method.resample()
methods.BARTMachineModel
to predict highest binary response level.BARTMachineModel
nu
parameter for numeric responses only.ModeledInput()
to SelectedModelFrame
, SelectedModelRecipe
, and TunedModelRecipe
.TunedInput()
.StackedModel
and SuperModel
training information.TreeModel
.ModeledInput()
and SelectedInput()
objects constructed with formulas and matrices.fit()
methods to ensure that unprepped recipes are passed to models, like TunedModed
, StackedModel
, SelectedModel
and SuperModel
, needing to replicate preprocessing steps in their resampling routines.GLMModel
to factor and matrix responses.fun
instead of deprecated fun.y
in ggplot2 functions.metricinfo()
results for factor responses.SplitControl()
to train on the split sample instead of the full dataset.fit()
formula and matrix methods are called with meta-models.print()
argument n
to data frame and matrix columns for more concise display of large data structures.step_kmeans()
, step_kmedoids()
, and step_spca()
.MLModel
slot y
.ModelFrame
and ModelRecipe
columns (casenames)
to (names)
.ModelFrame
inheritance from data.frame
.Terms
S4 classes for ModelFrame
slot terms
.ModeledInput
, SelectedInput
and TunedInput
classes and methods.SelectedFormula()
, SelectedMatrix()
, SelectedModelFrame()
, SelectedRecipe()
, and TunedRecipe()
.tune()
.stat.Curves
to stat.Curve
.stat.Train
to stat.train
.SelectedModel
, StackedModel
, SuperModel
, and TunedModel
.SelectedRecipe
and TunedRecipe
.MLModel
trainbits
slot.stat.Tune
to stat.Train
.SelectedFormula()
, SelectedMatrix()
, and SelectedModelFrame()
.BinomialMatrix
→ BinomialVariate
, DiscreteVector
→ DiscreteVariate
, NegBinomialVector
→ NegBinomialVariate
, and PoissonVector
→ PoissonVariate
.require
for user-specified packages to load during parallel execution of resampling algorithms.case_strata
to case_stratum
.object
argument to data
in ConfusionMatrix()
, SurvEvents()
, and SurvProbs()
.c
methods for BinomialVariate
, DiscreteVariate
, ListOf
, and SurvMatrix
.role_binom()
, role_case()
, role_surv()
, and role_term()
to set recipe roles.base
argument to varimp()
for log-transformed p-values.ParamSet
to ParameterGrid
.reset
global settings individually.as.data.frame
methods for Performance
, Performance
summary, PerformanceDiff
, PerformanceDiffTest
, and Resamples
.DiscreteVector
class and subclasses BinomialVector
, NegBinomialVector
, and PoissonVector
for discrete response variables.DiscreteVector
classes as follows.DiscreteVector
: all models applicable to numeric responses.BinomialVector
/NegBinomialVector
/PoissonVector
: BlackBoostModel
, GAMBoostModel
, GLMBoostModel
, GLMModel
, and GLMStepAICModel
.BinomialVector
/PoissonVector
: GLMNetModel
.PoissonVector
: GBMModel
and XGBModel
MLModel
.Calibration()
, Confusion()
, Curves()
, Lift()
, and Resamples()
with c
methods.Confusion
S3 class as ConfusionList
S4 class.metricinfo()
and modelinfo()
.expand.model()
.tune()
.metricinfo()
and modelinfo()
.ParamSet()
.as.MLModel()
for coercing MLModelFit
to MLModel
.tune()
; call fit()
with a SelectedModel
or TunedModel
instead.CVOptimismControl
).BootOptimismControl
error with 2D responses.max.print
for the number of models and data frame rows to show with print methods.SelectedRecipe()
.tune()
methods.MLModelFit
element fitbits
(MLFitBits
object) with mlmodel
(MLModel
object).VarImp
slot center
to shift
.expand_model()
, expand_params()
, and expand_steps()
.TunedRecipe()
.expand_model()
for model expansion over tuning parameters.expand_params()
for model parameters expansion.expand_steps()
for recipe step parameters expansion.MLModelFunction
and MLModelList
classes.MLModel
, MLModelFunction
, and MLModelList
.NNetModel
fit error with binary and factor responses.modelinfo()
function not found error.tune()
resampling failures.types
and design
arguments from MLModel()
.metricinfo()
and modelinfo()
.SelectedModel
.maximize
argument from tune()
and TunedModel
.StackedModel()
and SuperModel
.expand.model()
.KNNModel
tuning grid.TunedModel
.na.action
argument from ModelFrame
methods.MLModel()
argument types
to response_types
.MLModel()
argument design
to predictor_encoding
.expand.model()
to expand_model()
.BootOptimismControl
).ModelFrame
and ModelRecipe
and save to Resamples
.BinaryConfusionMatrix
and OrderedConfusionMatrix
classes.ConfusionMatrix
constructor.metricinfo()
to confusion matrices.Resamples
.ModelFrame
formulas.ModelFrame
response in first column.response
formula method.ICHomes
dataset.center
and scale
slot to VarImp
.ModelFrame
formulas.response
function argument from data
to newdata
.fit
, resample
, and tune
methods for design matrices.ModelFrame()
argument na.action
to na.rm
."exponential"
, "rayleigh"
, "weibull"
) estimation of baseline survival functions."weibull"
as the default distribution for survival mean estimation.Resamples
.na.rm
argument to calibration()
, confusion()
, performance()
, and performance_curve()
.span
argument to calibration()
.SurvMatrix
from S4 to S3 class.method
option to predict()
for Breslow, Efron (default), or Fleming-Harrington estimation of survival curves for Cox proportional hazards-based models.dist
option to predict()
for exponential or Weibull approximation to estimated survival curves.dist
option to calibration()
for distributional estimation of observed mean survival.dist
option to r2()
for distributional estimation of the total sum of squares mean.metricinfo()
and modelinfo()
.auc
, fnr
, fpr
, rpp
, tnr
, tpr
.SurvMatrix
classes for predicted survival events and probabilities to eliminate need for separate times
arguments in calibration, confusion, metrics, and performance functions.MLControl
argument surv_times
to times
.case_weight
and case_strata
variables.BARTModel
.accuracy
, f_score
, kappa2
, npv
, ppv
, pr_auc
, precision
, recall
, roc_index
, sensitivity
, specificity
cindex
, gini
, mae
, mse
, msle
, r2
, rmse
, rmsle
.performance
and metric methods for ConfusionMatrix
.MLModel
slot and constructor argument nvars
with design
.BARTMachineModel
, LARSModel
.gini
, multi-class pr_auc
and roc_auc
, multivariate rmse
, msle
, rmsle
.MLMetric
class for performance metrics.as.data.frame
method for ModelFrame
.expand.model
function.label
slot to MLModel
.metricinfo/modelinfo
support for mixed argument types.calibration
argument n
to breaks
.modelmetrics
function to performance
.ModelMetrics/Diff
classes to Performance/Diff
.MLModelTune
slot resamples
to performance
.AdaBagModel
, AdaBoostModel
, BlackBoostModel
, EarthModel
, FDAModel
, GAMBoostModel
, GLMBoostModel
, MDAModel
, NaiveBayesModel
, PDAModel
, RangerModel
, RPartModel
, TreeModel
modelmetrics
function.accuracy
, brier
, cindex
, cross_entropy
, f_score
, kappa2
, mae
, mse
, npv
, ppv
, pr_auc
, precision
, r2
, recall
, roc_auc
, roc_index
, sensitivity
, specificity
, weighted_kappa2
.cutoff
argument to confusion
function.modelinfo
and metricinfo
functions.modelmetrics
method for Resamples
.ModelMetrics
class with print
and summary
methods.response
method for recipe
.Calibration
constructor.Confusion
constructor.Lift
constructor.calibration
arguments to observed and predicted responses.confusion
arguments to observed and predicted responses.lift
arguments to observed and predicted responses.metrics
and stats
function arguments to accept function names.Resamples
to arguments with multiple models.CoxModel
, GLMModel
, and SurvRegModel
constructor definitions so that model control parameters are specified directly instead of with a separate control
argument/structure.predict(..., times = numeric())
function calls to survival model fits to return predicted values in the same direction as survival times.predict(..., times = numeric())
function calls to CForestModel
fits to return predicted means instead of medians.tune
function argument metrics
to be defined in terms of a user-specified metric or metrics.cutoff
, cutoff_index
, na.rm
, and summary
.LMModel
), linear discriminant analysis (LDAModel
), and quadratic discriminant analysis (QDAModel
).strata
argument of ModelFrame
or the role of "case_strata"
for recipe variables."case_weight"
for recipe variables.prepper
due to its relocation from rsample
to recipes
.KNNModel
), stacked regression models (StackedModel
), super learner models (SuperModel
), and extreme gradient boosting (XGBModel
).TrainControl
) and split training and test sets (SplitControl
).ModelFrame
class for general model formula and dataset specification.modelmetrics()
.predict()
to automatically preprocess recipes and to use training data as the newdata
default.tune()
to lists of models.summary()
argument stats
to functions.GBMModel
and GLMNetModel
.MLControl
argument na.rm
default from FALSE
to TRUE
.na.rm
argument from modelmetrics()
.Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.