Releases: brian-j-smith/MachineShop
Releases · brian-j-smith/MachineShop
MachineShop 3.7.0
Version Updates
3.7.0
- Compatibility updates for parsnip.
- Enable resampling by a grouping variable with
BootControl
,OOBControl
, andSplitControl
. - Enable resampling by a stratification variable with
SplitControl
. - Require R 4.1.0 or later.
MachineShop 3.6.2
Version Updates
3.6.2
- Add backward compatibility for older
MLModel
objects without ana.rm
slot. - Fix CRAN check warning: S3 generic/method consistency.
- Update
role_binom()
,role_case()
, androle_surv()
to remove the requirement that their variables be present innewdata
supplied topredict()
.
MachineShop 3.6.1
Version Updates
3.6.1
- Compatibility updates for ggplot2, Matrix, and recipes package dependencies.
MachineShop 3.6.0
Version Updates
3.6.0
- Add argument
na.rm
toMLModel()
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 thena.rm
values in suppliedMLModels
to automatically remove cases with missing values if not supported by their model fitting and prediction functions. - Add argument
prob.model
toSVMModel()
. - Add argument
verbose
tofit()
andpredict()
. - Fix
Error in as.data.frame(x) : object 'x' not found
issue when fitting aBARTMachineModel
that started occurring withbartMachine
package version 1.2.7. - Remove expired deprecations of
ModeledInput
andrpp()
. - Internal changes
- Add slot
na.rm
toMLModel
.
- Add slot
MachineShop 3.5.0
Version Updates
3.5.0
- Add argument
method
tor2()
for calculation of Pearson or Spearman correlation. - Add
predict()
S4 method forMLModelFit
. - Export
MLModelFunction()
. - Export
as.MLInput()
methods forMLModelFit
andModelSpecification
. - Export
as.MLModel()
method forModelSpecification
. - Improve recursive feature elimination of
SelectedInput
terms. - Improve speed of
StackedModel
andSuperModel
. - Internal changes
- Add
.MachineShop
list attribute toMLModelFit
. - Move field
mlmodel
inMLModelFit
tomodel
in.MachineShop
. - Move slot
input
inMLModel
to.MachineShop
. - Pass
.MachineShop
to thepredict
andvarimp
slot functions ofMLModel
.
- Add
MachineShop 3.4.3
Version Updates
3.4.3
- Fix
TypeError
independence()
with numeric dummy variables from recipes. - Prep
ModelRecipe
withretain = TRUE
for recipe steps that are skipped, for example, when test datasets are created. - Add generalized area under performance curves to
auc()
,pr_auc()
, androc_auc()
for multiclass factor responses.
MachineShop 3.4.2
Version Updates
3.4.2
- Add argument
select
torfe()
. - Fix object
perf_stats
not found inoptim()
.
MachineShop 3.4.1
Version Updates
3.4.1
- Add argument
conf
toset_optim_bayes()
. - Enable global grid expansion and tuning of
StackedModel
andSuperModel
inModelSpecification()
.
3.4.0
- Fixes
- Enable prediction with survival times of 0.
- Implement class
SelectedModelSpecification
. - Internal changes
- Deprecate classes
ModeledInput
,ModeledFrame
, andModeledRecipe
. - Remove unused class
TunedModeledRecipe
.
- Deprecate classes
- Expire deprecations
- Remove argument
fixed
fromTunedModel()
. - Remove
Grid()
.
- Remove argument
- Rename
rpp()
toppr()
. - Replace
ModeledInput()
withModelSpecification()
. - Require R >= 4.0.0.
- Use Olden algorithm for
NNetModel
model-specific variable importance.
MachineShop 3.3.0
Version Updates
3.3.0
- Add argument
.type
with options"glance"
and"tidy"
tosummary.MLModelFit()
. - Add case components data (stratification and grouping variables) to
print.Resample()
. - Add class and methods for
ModelSpecification
. - Add training parameters set functions
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 annealing
- Add
performance()
method forMLModel
to replicate the previous behavior ofsummary.MLModel()
. - Add
performance()
,plot()
, andsummary()
methods forTrainingStep
. - Add support for unordered plots of
Resample
performances. - Changes to argument
type
ofpredict()
.- Add option
"default"
for model-specific default predictions. - Add option
"numeric"
for numeric predictions. - Change option
"prob"
to be for probabilities between 0 and 1.
- Add option
- Change
confusion()
default behavior to convert factor probabilities to levels. - Rename argument
control
toobject
in set functions. - Rename argument
f
tofun
inroc_index()
. - Return a
ListOf
training step summaries fromsummary.MLModel()
. - Return a
TrainingStep
object fromrfe()
. - Support tibble-convertible objects as arguments to
expand_params()
. - Internal changes
- Add class
EnsembleModel
. - Add classes
MLOptimization
,GridSearch
,NullOptimization
,RandomGridSearch
, andSequentialOptimization
. - Add class
NullControl
. - Add slot
control
toPerformanceCurve
. - Add slot
method
toTrainingStep
. - Add slot
optim
toTrainingParams
. - Add slot
params
toMLInput
. - Inherit class
SelectedModel
fromEnsembleModel
. - Inherit class
StackedModel
fromEnsembleModel
. - Inherit class
SuperModel
fromStackedModel
. - Rename slot
case_comps
tovars
inResample
. - Rename slot
grid
tolog
inTrainingStep
.
- Add class
- Fixes
- error predicting single factor response in
GLMModel
- 'size(x@performance, 3)' error in
print.TrainingStep()
- 'Unmatched tuning parameters' error in
TunedModel()
- error predicting single factor response in
3.2.1
- Fix 'data' argument of wrong type error in
terms.formula()
. - Require >= 3.1.0 version of cli package.
MachineShop 3.2.0
Version Updates
3.2.0
- Add argument
distr
andmethod
todependence()
. - Add function
ParsnipModel()
for model specifications (model_spec
) from the parsnip package. - Add function
rfe()
for recursive feature elimination. - Add method
as.MLModel()
formodel_spec
andModeledInput
. - Add support for any model specification whose object has an
as.MLModel()
method. - Add support for cross-validation with case groups.
- Add support for names in argument
metric
ofauc()
. - Change argument
method
default from"model"
to"permute"
invarimp()
. - Change class
ModelFrame
to an S4 class; generally requires explicit conversion to a data frame withas.data.frame()
inMLModel
fit
andpredict
functions. - Change progress bar display from elapsed to estimated completion time.
- Changes to global settings
- Rename
stat.Trained
tostat.TrainingParams
. - Remove
stats.VarImp
.
- Rename
- Changes to internal classes
- Add class
ParsnipModel
. - Add class
SurvTimes
. - Add class
TrainingParams
. - Add class union
Grid
. - Add class union
Params
. - Add column
name
,selected
, andmetrics
to slotgrid
ofTrainingStep
class. - Add slot
grid
toTunedInput
. - Add slot
id
toMLInput
andMLModel
classes. - Add slot
id
andname
toTrainingStep
class. - Add slot
models
toSelectedModel
. - Remove slot
name
fromMLControl
classes. - Remove slot
selected
,values
, andmetric
fromTrainingStep
class. - Remove slot
shift
fromVariableImportance
class. - Rename class
Grid
toTuningGrid
. - Rename class
Resamples
toResample
. - Rename class
TrainStep
toTrainingStep
. - Rename class
VarImp
toVariableImportance
. - Rename classes of
MLControl
.MLBootControl
→BootControl
MLBootOptimismControl
→BootOptimismControl
MLCVControl
→CVControl
MLCVOptimismControl
→CVOptimismControl
MLOOBControl
→OOBControl
MLSplitControl
→SplitControl
MLTrainControl
→TrainControl
- Rename column
Input
andModel
toparams
in slotgrid
ofTrainingStep
class. - Rename column
Resample
toIteration
inResample
class - Rename slot
x
toinput
inMLModel
class.
- Add class
- Changes to
XGBModel
- Change argument default for
nrounds
from 1 to 100. - Rearrange constructor arguments.
- Reduce number of tuning grid parameters
- Include
nrounds
andmax_depth
in automated grids forXGBDARTModel
andXGBTreeModel
. - Include
nrounds
,lambda
, andalpha
in automated grid forXGBLinearModel
.
- Include
- Compute survival probabilities for
survival:aft
prediction. - Change default survival objective from
survival:cox
tosurvival:aft
.
- Change argument default for
- Format and condense printout of objects.
- Include all computed performance metrics in
TrainingStep
objects and output. - Remove shift from variable importance scaling in
varimp()
. - Rename and redefine dispatch (first) arguments in functions.
model
→object
inTunedModel()
x
→object
inexpand_model()
x
→formula
/input
/model
inexpand_modelgrid()
,fit()
,ModelFrame()
,resample()
,rfe()
methodsx
→formula
/object
/model
inModeledInput()
methodsx
→object
inParameterGrid()
methodsx
→control
inset_monitor()
,set_predict()
,set_strata()
x
→object
inTunedInput()
- Rename function
Grid()
toTuningGrid()
. - Reorder optional arguments in
ModelFrame()
. - Save model constructor arguments as the list elements in
MLModel
params
slots.