Composable Preprocessing Operators

Description

mlrCPO is a toolset that enriches mlr with a diverse set of preprocessing operators. Composable Preprocessing Operators (“CPO”s) are first-class R objects that can be applied to data.frames and mlr Tasks to modify data, they can be attached to mlr Learners to add preprocessing to machine learning algorithms, and they can be composed to form preprocessing pipelines.

mlrCPO focuses on preprocessing as part of automated machine learning pipelines. This means that it is designed with the expectation that the same preprocessing options are applied to incoming training data, and test data. A common mistake in machine learning is that a machine learning method is evaluated (e.g. using resampling) on a dataset after that dataset has been cleaned up and preprocessed in one go. The proper evaluation would need to consider that the preprocessing of training data may not be influenced by any information contained in the test data set. mlrCPO takes this duality into account by providing CPO objects that run on training data, and which then create CPOTrained objects that can be used on test data (or entirely new prediction data).

This focus on preprocessing is the reason for a strict separation between “Feature Operation” CPOs, “Target Operation” CPOs, and “Retrafoless” CPOs (see OperatingType). The first class only changes (predictor) features of a dataset, and does so in a way reproducible on test data. The second class only changes (outcome) target data of a dataset, and is then able to invert the prediction, made by a learner on new data, back to the space of the original target data. The “Retrafoless” CPO only operates during training and may only add or subtract data rows (e.g. for SMOTE-ing or subsampling), without transforming the space of either predictor or outcome variables.

CPO's design is supposed to help its user avoid bugs and errors. Therefore it often avoids doing things implicitly and relies on explicit commands e.g. for removing data or converting between datatypes. It has certain restrictions in place (e.g. CPOProperties, CPOTrainedCapability) that try to make it hard to do the wrong thing while not being in the way of the right thing.

Other packages with similar, partially overlapping functionality are recipes, dplyr, and caret.

CPO Composition / Attachment / Application Operator

Description

This operator “pipes” data from the source into the sink object.

If both objects are a CPO object, or both are a CPOTrained object, they will be composed. A new object, representing the operation of performing both object's operations in succession, will be created, which can be handled like a new CPO or CPOTrained object. See composeCPO.

If the source object is a data.frame or a link[mlr]{Task}, the transformation operation will be applied to this data, and the same resulting data will be returned. See applyCPO.

If the sink object is a Learner, the CPO will be attached to this learner. The same operation will be performed during the train and predict phase; the behaviour during the predict phase may furthermore be depend on the training data. See attachCPO.

Note that you can not link a data.frame or Task directly to a Learner, since this operation is not algebraically associative with the composition of CPOs. Use train for this.

The %<<% operator is synonymous with %>>% with source and sink argument swapped.

The %>|% and %|<% operators perform piping followed by application of retrafo. The %>|% evaluates the expression to its right before the expression to its left, so it may be used in the most natural way without parentheses:

data %>|% cpo1 %>>% cpo2

is the same as

retrafo(data %>>% cpo1 %>>% cpo2).

The %<>>% and %<<<% operators perform the piping operation and assign the result to the left hand variable. This way it is possible to apply a CPO, or to attach a CPO to a Learner, and just keep the resulting object. The assignment operators evaluate their right hand side before their left hand side, so it is possible to build long chains that end up writing to the leftmost variable. Therefore the expression

data %<>>% cpo1 %<>>% cpo2 %>>% cpo3

is the same as

cpo1 = cpo1 %>>% cpo2 %>>% cpo3
data = data %>>% cpo1

Usage

cpo1 %>>% cpo2

cpo2 %<<% cpo1

cpo1 %<>>% cpo2

cpo2 %<<<% cpo1

cpo1 %>|% cpo2

cpo2 %|<% cpo1

Arguments

Value

[data.frame | Task | CPO | CPOTrained].

See Also

Other operators: CPO, applyCPO(), as.list.CPO, attachCPO(), composeCPO(), pipeCPO()

Other retrafo related: CPOTrained, NULLCPO, applyCPO(), as.list.CPO, clearRI(), getCPOClass(), getCPOName(), getCPOOperatingType(), getCPOPredictType(), getCPOProperties(), getCPOTrainedCPO(), getCPOTrainedCapability(), getCPOTrainedState(), is.retrafo(), makeCPOTrainedFromState(), pipeCPO(), print.CPOConstructor()

Other inverter related: CPOTrained, NULLCPO, applyCPO(), as.list.CPO, clearRI(), getCPOClass(), getCPOName(), getCPOOperatingType(), getCPOPredictType(), getCPOProperties(), getCPOTrainedCPO(), getCPOTrainedCapability(), getCPOTrainedState(), is.inverter(), makeCPOTrainedFromState(), pipeCPO(), print.CPOConstructor()

Other CPO lifecycle related: CPO, CPOConstructor, CPOLearner, CPOTrained, NULLCPO, attachCPO(), composeCPO(), getCPOClass(), getCPOConstructor(), getCPOTrainedCPO(), identicalCPO(), makeCPO()

Examples

# PCA-rotate pid.task
rotated.pid.task = pid.task %>>% cpoScale() %>>% cpoPca()

# Centering / Scaling *after* PCA
newPCA = cpoPca() %>>% cpoScale()

# Attach the above to learner
pcaLogreg = newPCA %>>% makeLearner("classif.logreg")

# append cpoAsNumeric to newPCA
newPCA %<>>% cpoAsNumeric()
print(newPCA)

# prepend cpoAsNumeric to pcaLogreg
pcaLogreg %<<<% cpoAsNumeric()

Composable Preprocessing Operators

Description

Composable Preprocessing Operators, or CPO, are the central entity provided by the mlrCPO package. CPOs can perform operations on a data.frame or a Task, for the latter even modifying target values and converting between different Task types.

CPOs can be “composed” using the %>>% operator, the composeCPO function, or the pipeCPO function, to create new (“compound”) operators that perform multiple operations in a pipeline. While all CPOs have the class “CPO”, primitive (i.e. not compound) CPOs have the additional class “CPOPrimitive”, and compound CPOs have the class “CPOPipeline”. It is possible to split a compound CPOs into its primitive constituents using as.list.CPO.

CPOs can be “attached” to a mlr-Learner objects to create CPOLearners, using the %>>% operator, or the attachCPO function. These CPOLearners fit the model specified by the Learner to the data after applying the attached CPO. Many CPOs can be attached to a Learner sequentially, or in form of a compound CPO.

CPOs can be “applied” to a data.frame or a Task using the %>>% operator, or the applyCPO function. Applying a CPO performs the operations specified by the (possibly compound) CPO, and returns the modified data. This data also contains a “retrafo” and and “inverter” tag, which can be accessed using the retrafo and inverter functions to get CPORetrafo and CPOInverter objects, respectively. These objects represent the “trained” CPOs that can be used when performing validation or predictions with new data.

Hyperparameters

CPOs can have hyperparameters that determine how they operate on data. These hyperparameters can be set during construction, as function parameters of the CPOConstructor, or they can potentially be modified later as exported hyperparameters. Which hyperparameters are exported is controlled using the export parameter of the CPOConstructor when the CPO was created. Hyperparameters can be listed using getParamSet, queried using getHyperPars and set using setHyperPars.

S3 properties

A CPO object should be treated as an opaque object and should only be queried / modified using the given set* and get* functions. A list of them is given below in the section “See Also”–“cpo-operations”.

Special CPO

A special CPO is NULLCPO, which functions as the neutral element of the %>>% operator and represents the identity operation on data.

See Also

print.CPO for possibly verbose printing.

Other CPO lifecycle related: CPOConstructor, CPOLearner, CPOTrained, NULLCPO, %>>%(), attachCPO(), composeCPO(), getCPOClass(), getCPOConstructor(), getCPOTrainedCPO(), identicalCPO(), makeCPO()

Other operators: %>>%(), applyCPO(), as.list.CPO, attachCPO(), composeCPO(), pipeCPO()

Other getters and setters: getCPOAffect(), getCPOClass(), getCPOConstructor(), getCPOId(), getCPOName(), getCPOOperatingType(), getCPOPredictType(), getCPOProperties(), getCPOTrainedCPO(), getCPOTrainedCapability(), setCPOId()

Other CPO classifications: getCPOClass(), getCPOOperatingType(), getCPOTrainedCapability()

Examples

class(cpoPca())  # c("CPOPrimitive", "CPO")
class(cpoPca() %>>% cpoScale())  # c("CPOPipeline", "CPO")
print(cpoPca() %>>% cpoScale(), verbose = TRUE)

getHyperPars(cpoScale(center = FALSE))

head(getTaskData(iris.task %>>% cpoScale()))

Constructor for CPO Objects

Description

CPO objects are created by calling CPOConstructors, which are R functions that have some parameters in common, use a convenient print.CPOConstructor generic, and always return a CPO object. The mlrCPO package provides many CPOConstructor functions, which can be listed using listCPO. It is also possible to create custom CPOConstructors using makeCPO, makeCPORetrafoless, link{makeCPOTargetOp}, and makeCPOExtendedTrafo.

Arguments

Value

[CPO] the constructed CPO.

CPO creation

CPOConstructors can be called like any R function, with any parameters given. Besides parameters that are common to most CPOConstructors (listed below), it is possible to set CPO-specific hyperparameters in the construction. Parameters that are being exported can also be modified later using the CPO object, see the documentation there.

affect.* parameters

When creating a CPO, it is possible to choose which columns of the given data the CPO operates on, and which columns it will ignore. This is done using the affect.* parameters. It is possible to choose columns by types, indices, names, or a regular expression matching names.

See Also

print.CPOConstructor for possibly verbose printing.

Other CPO lifecycle related: CPO, CPOLearner, CPOTrained, NULLCPO, %>>%(), attachCPO(), composeCPO(), getCPOClass(), getCPOConstructor(), getCPOTrainedCPO(), identicalCPO(), makeCPO()

Other CPOConstructor related: getCPOClass(), getCPOConstructor(), getCPOName(), identicalCPO(), makeCPO(), print.CPOConstructor()

Examples

class(cpoPca)  # c("CPOConstructor", "function")
print(cpoPca)  # default printer
print(cpoPca, verbose = TRUE)  # shows the trafo / retrafo functions

cpoPca()  # creating a CPO
class(cpoPca())  # c("CPOPrimitive", "CPO")

CPO Learner Object

Description

CPO Learners are created when a CPO gets attached to an mlr-Learner object. The resulting learner performs the operation described by the attached CPO before fitting the model specified by the Learner. It is possible to attach compound CPOs, and it is possible to attach more CPOs to a learner that is already a CPOLearner. If the attached CPO exports hyperparameters, these become part of the newly created learner and can be queried and set using functions such as getParamSet, getHyperPars, and setHyperPars.

The model created when training a CPOLearner also contains the relevant CPORetrafo information to be applied to prediction data; this can be retrieved using retrafo. The CPOInverter functionality is handled equally transparently by the model.

A CPOLearner can possibly have different LearnerProperties than the base Learner to which it is attached. This depends on the CPO's properties, see CPOProperties.

It is possible to retrieve the CPOLearner's base learner using getLearnerBare, and to get the attached CPOs using getLearnerCPO.

See Also

Other CPO lifecycle related: CPO, CPOConstructor, CPOTrained, NULLCPO, %>>%(), attachCPO(), composeCPO(), getCPOClass(), getCPOConstructor(), getCPOTrainedCPO(), identicalCPO(), makeCPO()

Other CPOLearner related: attachCPO(), getLearnerBare(), getLearnerCPO()

Examples

lrn = makeLearner("classif.logreg")
cpolrn = cpoScale() %>>% lrn
print(cpolrn)

getLearnerBare(cpolrn)  # classif.logreg Learner
getLearnerCPO(cpolrn)  # cpoScale() CPO

getParamSet(cpolrn)  # includes cpoScale hyperparameters

model = train(cpolrn, pid.task)  # behaves like a learner
retrafo(model)  # the CPORetrafo that was trained

predict(model, pid.task)  # otherwise behaves like an mlr model

Get the Retransformation or Inversion Function from a Resulting Object

Description

When applying a CPO to a data.frame or Task, the data is not only changed, additionally a retransformation and an inversion object is created that can be applied to other data of the same kind. This is useful if new data (for prediction or validation) is to be handled in the same machine learning procedure.

For example, when performing PCA on training data using cpoPca, the rotation matrix is saved and can be used on new (prediction) data. As another example, consider a log-transformation of the target column in a regression problem. When predictions are made with new data, it may be useful to invert the transformation on the predicted values by exponentiating them.

The information created when a CPO is applied is saved in a CPORetrafo object, and a CPOInverter object, which are both saved as attributes. The retrafo and inverter function retrieve these objects. It is furthermore possible to set these attributes using the retrafo<- and inverter<- functions, using constructs like retrafo(data) <- retr.obj. The retrafo or inverter attributes can be reset individually by setting them to NULL: retrafo(data) <- NULL, or by using the clearRI function.

When chaining %>>% on a data object, the retrafo and inverter associated with the result is also chained automatically. Beware, however, that this just accesses the retrafo attribute internally. Therefore, if you plan to do apply multiple transformations with other operations in between, make sure to reset the retrafo function by setting it to NULL, or using the clearRI function. See examples.

Usage

retrafo(data)

inverter(data)

retrafo(data) <- value

inverter(data) <- value

Arguments

Value

[CPOTrained]. The retransformation function that can be applied to new data. This is a CPORetrafo object for retrafo or a CPOInverter object for inverter.

CPORetrafo and CPOInverter

CPORetrafo and CPOInverter objects are members of the CPOTrained class, which can be handled similarly to CPO objects: Their hyperparameters can be inspected using getParamSet and link[mlr]{getHyperPars}, print.CPOTrained is used for (possibly verbose) printing. To apply the retrafo or inverter transformation represented by the object to data, use the applyCPO or %>>% function.

CPOTrained objects can be chained using %>>% or pipeCPO, and broken into primitives using as.list.CPOTrained. However, since the CPOTrained objects represent transformations that relate closely to the data used to train it (and therefore to the position within a CPO pipeline), it is only advisable to chain or break apart CPOTrained pipes for inspection, or if you really know what you are doing.

(Primitive) CPORetrafo objects can be inspected using getCPOTrainedState, and it is possible to create new CPORetrafo objects from (possibly modified) retrafo state using makeCPOTrainedFromState.

Difference between CPORetrafo and CPOInverter

The fundamental difference between CPORetrafo and CPOInverter is that a CPORetrafo is created only when a CPO is applied to a data set, and is used to perform the same transformation on new (prediction) data. The CPOInverter is created whenever a CPO or CPORetrafo is applied to data (whether training or prediction data). It is in fact used to invert the transformation done to the target column of a Task. Since this operation may depend on the new prediction data, and not only on the training data fed to the CPO when the CPORetrafo was created, the CPOInverter object is more closely bound to the particular data set used to create it.

In some cases a target transformation is independent of the data used to create it (e.g. log-transform of a regression target column); in that case the CPORetrafo can be used with invert. This is the concept of CPOTrainedCapability, which can be queried using getCPOTrainedCapability.

Using CPORetrafo

CPORetrafo objects can be applied to new data sets using the %>>% operator, the applyCPO generic, or the predict generic, all of which perform the same action.

Using CPOInverter

To use a CPOInverter, use the invert function.

See Also

clearRI about the problem of needing to reset retrafo and inverter attributes sometimes.

Other CPO lifecycle related: CPO, CPOConstructor, CPOLearner, NULLCPO, %>>%(), attachCPO(), composeCPO(), getCPOClass(), getCPOConstructor(), getCPOTrainedCPO(), identicalCPO(), makeCPO()

Other retrafo related: NULLCPO, %>>%(), applyCPO(), as.list.CPO, clearRI(), getCPOClass(), getCPOName(), getCPOOperatingType(), getCPOPredictType(), getCPOProperties(), getCPOTrainedCPO(), getCPOTrainedCapability(), getCPOTrainedState(), is.retrafo(), makeCPOTrainedFromState(), pipeCPO(), print.CPOConstructor()

Other inverter related: NULLCPO, %>>%(), applyCPO(), as.list.CPO, clearRI(), getCPOClass(), getCPOName(), getCPOOperatingType(), getCPOPredictType(), getCPOProperties(), getCPOTrainedCPO(), getCPOTrainedCapability(), getCPOTrainedState(), is.inverter(), makeCPOTrainedFromState(), pipeCPO(), print.CPOConstructor()

Examples

traindat = subsetTask(pid.task, 1:400)
preddat = subsetTask(pid.task, 401:768)

trained = traindat %>>% cpoPca()
reFun = retrafo(trained)
predicted = preddat %>>% reFun
head(getTaskData(predicted))

# chaining works
trained = traindat %>>% cpoPca() %>>% cpoScale()
reFun = retrafo(trained)
predicted = preddat %>>% reFun
head(getTaskData(predicted))

# reset the retrafo when doing other steps!

trained.tmp = traindat %>>% cpoPca()
reFun1 = retrafo(trained.tmp)

imp = impute(trained.tmp)
trained.tmp = imp$task  # nonsensical example
retrafo(trained.tmp) = NULL  # NECESSARY HERE

trained = trained.tmp %>>% cpoScale()

reFun2 = retrafo(trained)
predicted = getTaskData(reimpute(preddat %>>% reFun1, imp$desc),
  target.extra = TRUE)$data %>>% reFun2

CPO Composition Neutral Element

Description

NULLCPO is the neutral element of CPO and CPOTrained composition when using %>>% or composeCPO. It is furthermore no effect when attached to a Learner using attachCPO (or %>>%), or when applied to data using applyCPO, invert, or predict (or, again, %>>%).

NULLCPO works as a stand-in for certain operations that have an "empty" return value: It is returned when retrafo and inverter are applied to an object that has no retrafo or inverter associated with it, and by pipeCPO when applied to an empty list.

NULLCPO can be checked using is.nullcpo, and converted from or to NULL using nullToNullcpo and nullcpoToNull. Otherwise it behaves very similarly to other CPO or CPOTrained objects.

Usage

NULLCPO

Format

An object of class NULLCPO (inherits from CPOPrimitive, CPORetrafo, CPOInverter, CPOTrained, CPO) of length 0.

See Also

Other retrafo related: CPOTrained, %>>%(), applyCPO(), as.list.CPO, clearRI(), getCPOClass(), getCPOName(), getCPOOperatingType(), getCPOPredictType(), getCPOProperties(), getCPOTrainedCPO(), getCPOTrainedCapability(), getCPOTrainedState(), is.retrafo(), makeCPOTrainedFromState(), pipeCPO(), print.CPOConstructor()

Other inverter related: CPOTrained, %>>%(), applyCPO(), as.list.CPO, clearRI(), getCPOClass(), getCPOName(), getCPOOperatingType(), getCPOPredictType(), getCPOProperties(), getCPOTrainedCPO(), getCPOTrainedCapability(), getCPOTrainedState(), is.inverter(), makeCPOTrainedFromState(), pipeCPO(), print.CPOConstructor()

Other CPO lifecycle related: CPO, CPOConstructor, CPOLearner, CPOTrained, %>>%(), attachCPO(), composeCPO(), getCPOClass(), getCPOConstructor(), getCPOTrainedCPO(), identicalCPO(), makeCPO()

Other NULLCPO related: is.nullcpo(), nullToNullcpo(), nullcpoToNull()

Apply a CPO to Data

Description

The given transformation will be applied to the data in the given Task or data.frame.

If the input data is a data.frame, the returned object will in most cases also be a data.frame, with exceptions if the applied CPO performs a conversion to a Task. If the input data is a Task, its type will only be changed to a different type of Task if the applied CPO performs such a conversion.

The %>>% operator can be used synonymously to apply CPO objects to data. In case of CPORetrafo, predict can be used synonymously.

Usage

applyCPO(cpo, task)

Arguments

Value

[Task | data.frame]. The transformed data, augmented with a inverter and possibly a retrafo tag.

Application of CPO

Application of a CPO is supposed to perform preprocessing on a given data set, to prepare it e.g. for model fitting with a Learner, or for other data handling tasks. When this preprocessing is performed, care is taken to make the transformation repeatable on later prediction or validation data. For this, the returned data set will have a CPORetrafo and CPOInverter object attached to it, which can be retrieved using retrafo and inverter. These can be used to perform the same transformation on new data, or to invert a prediction made with the transformed data.

An applied CPO can change the content of feature columns, target columns of Tasks, and may even change the number of rows of a given data set.

Application of CPORetrafo

Application of a CPORetrafo is supposed to perform a transformation that mirrors the transformation done before on a training data set. It should be used when trying to make predictions from new data, using a model that was trained with data preprocessed using a CPO. The predictions made may then need to be inverted. For this, the returned data set will have a CPOInverter object attached to it, which can be retrieved using inverter.

An applied CPORetrafo may change the content of feature columns and target columns of Tasks, but will never change the number or order of rows of a given data set.

See Also

Other operators: CPO, %>>%(), as.list.CPO, attachCPO(), composeCPO(), pipeCPO()

Other retrafo related: CPOTrained, NULLCPO, %>>%(), as.list.CPO, clearRI(), getCPOClass(), getCPOName(), getCPOOperatingType(), getCPOPredictType(), getCPOProperties(), getCPOTrainedCPO(), getCPOTrainedCapability(), getCPOTrainedState(), is.retrafo(), makeCPOTrainedFromState(), pipeCPO(), print.CPOConstructor()

Other inverter related: CPOTrained, NULLCPO, %>>%(), as.list.CPO, clearRI(), getCPOClass(), getCPOName(), getCPOOperatingType(), getCPOPredictType(), getCPOProperties(), getCPOTrainedCPO(), getCPOTrainedCapability(), getCPOTrainedState(), is.inverter(), makeCPOTrainedFromState(), pipeCPO(), print.CPOConstructor()

Split a Pipeline into Its Constituents

Description

Split a compound CPO or CPOTrained into a list of its constituent parts.

This is useful for inspection of pipelines, or for possible rearrangements or changes of pipelines. The resulting list can be changed and rebuilt using pipeCPO.

Usage

## S3 method for class 'CPOPrimitive'
as.list(x, ...)

## S3 method for class 'CPOTrained'
as.list(x, ...)

Arguments

Value

[list of CPO | list of CPOTrained]. The primitive constituents of x.

See Also

Other operators: CPO, %>>%(), applyCPO(), attachCPO(), composeCPO(), pipeCPO()

Other retrafo related: CPOTrained, NULLCPO, %>>%(), applyCPO(), clearRI(), getCPOClass(), getCPOName(), getCPOOperatingType(), getCPOPredictType(), getCPOProperties(), getCPOTrainedCPO(), getCPOTrainedCapability(), getCPOTrainedState(), is.retrafo(), makeCPOTrainedFromState(), pipeCPO(), print.CPOConstructor()

Other inverter related: CPOTrained, NULLCPO, %>>%(), applyCPO(), clearRI(), getCPOClass(), getCPOName(), getCPOOperatingType(), getCPOPredictType(), getCPOProperties(), getCPOTrainedCPO(), getCPOTrainedCapability(), getCPOTrainedState(), is.inverter(), makeCPOTrainedFromState(), pipeCPO(), print.CPOConstructor()

Attach a CPO to a Learner

Description

A CPO object can be attached to a Learner object to create a pipeline combining preprocessing and model fitting. When the resulting CPOLearner is used to create a model using train, the attached CPO will be applied to the data before the internal model is trained. The resulting model will also contain the required CPOTrained elements, and apply the necessary CPORetrafo objects to new prediction data, and the CPOInverter objects to predictions made by the internal model.

The %>>% operator can be used synonymously to attach CPO objects to Learners.

Usage

attachCPO(cpo, learner)

Arguments

See Also

Other operators: CPO, %>>%(), applyCPO(), as.list.CPO, composeCPO(), pipeCPO()

Other CPO lifecycle related: CPO, CPOConstructor, CPOLearner, CPOTrained, NULLCPO, %>>%(), composeCPO(), getCPOClass(), getCPOConstructor(), getCPOTrainedCPO(), identicalCPO(), makeCPO()

Other CPOLearner related: CPOLearner, getLearnerBare(), getLearnerCPO()

Clear Retrafo and Inverter Attributes

Description

When applying CPOs to data, the operation entails saving the CPOTrained information that gets generated to an attribute of the resulting object. This is a useful solution to the problem that applying multiple CPOs should also lead to a retrafo object that performs the same multiple operations. However, sometimes this may lead to surprising and unwanted results when a CPO is applied and not meant to be part of a trafo-retrafo machine learning pipeline, e.g. for dropping columns that occur in training but not in prediction data. In that case, it is necessary to reset the retrafo and possibly inverter attributes of the data being used. This can be done either by using retrafo(data) <- NULL, or by using clearRI. clearRI clears both retrafo and inverter attributes.

Usage

clearRI(data)

Arguments

Value

[data.frame | Task | WrappedModel] the data after stripping all retrafo and inverter attributes.

See Also

Other retrafo related: CPOTrained, NULLCPO, %>>%(), applyCPO(), as.list.CPO, getCPOClass(), getCPOName(), getCPOOperatingType(), getCPOPredictType(), getCPOProperties(), getCPOTrainedCPO(), getCPOTrainedCapability(), getCPOTrainedState(), is.retrafo(), makeCPOTrainedFromState(), pipeCPO(), print.CPOConstructor()

Other inverter related: CPOTrained, NULLCPO, %>>%(), applyCPO(), as.list.CPO, getCPOClass(), getCPOName(), getCPOOperatingType(), getCPOPredictType(), getCPOProperties(), getCPOTrainedCPO(), getCPOTrainedCapability(), getCPOTrainedState(), is.inverter(), makeCPOTrainedFromState(), pipeCPO(), print.CPOConstructor()

Examples

# without clearRI
transformed = iris.task %>>% cpoPca()
transformed2 = transformed %>>% cpoScale()
retrafo(transformed2)  # [RETRAFO pca]=>[RETRAFO scale]

transformed = iris.task %>>% cpoPca()
transformed2 = clearRI(transformed) %>>% cpoScale()
retrafo(transformed2)  # [RETRAFO scale]

CPO Composition

Description

Composes CPO or CPOTrained objects. The %>>% operator can be used synonymously to compose CPO objects.

Composition of operators is one of the main features they provide: this makes it possible for complex operations to be represented by single objects. Compound operators represent the operation of applying both its constituent operations in succession. Compound operators can themselves be composed to form arbitrarily long chains of operators.

Compound objects behave, in most ways, like primitive objects. Some exceptions are:

• Compound CPOs do not have an ID, so getCPOId and setCPOId will not work on them.

• Compound CPOs have no 'affect' property, so getCPOAffect will not work.

While CPOTrained operators can be composed just as CPO operators, this is only recommended in cases where the same primitive CPOTrained objects where retrieved using as.list.CPOTrained. This is because CPOTrained are closely related to the data that was used to create it, and therefore on their original position in the CPO pipeline during training.

Usage

composeCPO(cpo1, cpo2)

Arguments

Value

[CPO | CPOTrained]. The operation representing the application of cpo1 followed by the application of cpo2.

See Also

Other operators: CPO, %>>%(), applyCPO(), as.list.CPO, attachCPO(), pipeCPO()

Other CPO lifecycle related: CPO, CPOConstructor, CPOLearner, CPOTrained, NULLCPO, %>>%(), attachCPO(), getCPOClass(), getCPOConstructor(), getCPOTrainedCPO(), identicalCPO(), makeCPO()

Add 'covr' coverage to CPOs

Description

Use this if you want to check code coverage of CPOs using covr. The functions inside CPOs is originally not accessible to covr, so covrTraceCPOs needs to be called in the .onAttach function. Note that putting it in .onLoad will not work.

Currently, for this to work, the mb706 fork of covr needs to be used. To install it, call

devtools::install_github("mb706/covr")

To use it on Travis CI, add the line - mb706/covr under the r_github_packages: category.

This function comes at no runtime penalty: If the R_COVR environment variable is not set to “true”, then it only has an effect if force is TRUE.

Usage

covrTraceCPOs(env = parent.env(parent.frame()), force = FALSE)

Arguments

Value

[invisible(NULL)].

Apply a Function Element-Wise

Description

This is a CPOConstructor to be used to create a CPO. It is called like any R function and returns the created CPO.

The function must either vectorize over the given data, or will be applied to each data element on its own.

It must not change the type of the data, i.e. numeric data must remain numeric etc.

If the function can only handle a subset of the given columns, e.g. only a certain type, use affect.* arguments.

Usage

cpoApplyFun(
  fun,
  param = NULL,
  vectorize = TRUE,
  make.factors = TRUE,
  id,
  export = "export.default",
  affect.type = NULL,
  affect.index = integer(0),
  affect.names = character(0),
  affect.pattern = NULL,
  affect.invert = FALSE,
  affect.pattern.ignore.case = FALSE,
  affect.pattern.perl = FALSE,
  affect.pattern.fixed = FALSE
)

Arguments

Value

[CPO].

CPOTrained State

The created state is empty.

General CPO info

This function creates a CPO object, which can be applied to Tasks, data.frames, link[mlr]{Learner}s and other CPO objects using the %>>% operator.

The parameters of this object can be changed after creation using the function setHyperPars. The other hyper-parameter manipulating functins, getHyperPars and getParamSet similarly work as one expects.

If the “id” parameter is given, the hyperparameters will have this id as aprefix; this will, however, not change the parameters of the creator function.

Calling a CPOConstructor

CPO constructor functions are called with optional values of parameters, and additional “special” optional values. The special optional values are the id parameter, and the affect.* parameters. The affect.* parameters enable the user to control which subset of a given dataset is affected. If no affect.* parameters are given, all data features are affected by default.

See Also

Other CPOs: cpoApplyFunRegrTarget(), cpoAsNumeric(), cpoCache(), cpoCbind(), cpoCollapseFact(), cpoDropConstants(), cpoDropMostlyConstants(), cpoDummyEncode(), cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), cpoFixFactors(), cpoIca(), cpoImpactEncodeClassif(), cpoImpactEncodeRegr(), cpoImpute(), cpoImputeConstant(), cpoImputeHist(), cpoImputeLearner(), cpoImputeMax(), cpoImputeMean(), cpoImputeMedian(), cpoImputeMin(), cpoImputeMode(), cpoImputeNormal(), cpoImputeUniform(), cpoLogTrafoRegr(), cpoMakeCols(), cpoMissingIndicators(), cpoModelMatrix(), cpoOversample(), cpoPca(), cpoProbEncode(), cpoQuantileBinNumerics(), cpoRegrResiduals(), cpoResponseFromSE(), cpoSample(), cpoScale(), cpoScaleMaxAbs(), cpoScaleRange(), cpoSelect(), cpoSmote(), cpoSpatialSign(), cpoTransformParams(), cpoWrap(), makeCPOCase(), makeCPOMultiplex()

Transform a Regression Target Variable

Description

This is a CPOConstructor to be used to create a CPO. It is called like any R function and returns the created CPO.

Apply a given function to the target column of a regression Task.

Usage

cpoApplyFunRegrTarget(
  trafo,
  invert.response = NULL,
  invert.se = NULL,
  param = NULL,
  vectorize = TRUE,
  gauss.points = 23,
  id,
  export = "export.default",
  affect.type = NULL,
  affect.index = integer(0),
  affect.names = character(0),
  affect.pattern = NULL,
  affect.invert = FALSE,
  affect.pattern.ignore.case = FALSE,
  affect.pattern.perl = FALSE,
  affect.pattern.fixed = FALSE
)

Arguments

Value

[CPO].

Details

When both mean and se prediction is available, it may be possible to make more accurate mean inversion than for the response predict.type, using integrals or approximations like the delta method. In such cases it may be advisable to prepend this CPO with the cpoResponseFromSE CPO.

Note when trafo or invert.response take more than one argument, the second argument will be set to the value of param. This may lead to unexpected results when using functions with rarely used parameters, e.g. log. In these cases, it may be necessary to wrap the function: trafo = function(x) log(x).

General CPO info

This function creates a CPO object, which can be applied to Tasks, data.frames, link[mlr]{Learner}s and other CPO objects using the %>>% operator.

The parameters of this object can be changed after creation using the function setHyperPars. The other hyper-parameter manipulating functins, getHyperPars and getParamSet similarly work as one expects.

If the “id” parameter is given, the hyperparameters will have this id as aprefix; this will, however, not change the parameters of the creator function.

Calling a CPOConstructor

CPO constructor functions are called with optional values of parameters, and additional “special” optional values. The special optional values are the id parameter, and the affect.* parameters. The affect.* parameters enable the user to control which subset of a given dataset is affected. If no affect.* parameters are given, all data features are affected by default.

See Also

Other CPOs: cpoApplyFun(), cpoAsNumeric(), cpoCache(), cpoCbind(), cpoCollapseFact(), cpoDropConstants(), cpoDropMostlyConstants(), cpoDummyEncode(), cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), cpoFixFactors(), cpoIca(), cpoImpactEncodeClassif(), cpoImpactEncodeRegr(), cpoImpute(), cpoImputeConstant(), cpoImputeHist(), cpoImputeLearner(), cpoImputeMax(), cpoImputeMean(), cpoImputeMedian(), cpoImputeMin(), cpoImputeMode(), cpoImputeNormal(), cpoImputeUniform(), cpoLogTrafoRegr(), cpoMakeCols(), cpoMissingIndicators(), cpoModelMatrix(), cpoOversample(), cpoPca(), cpoProbEncode(), cpoQuantileBinNumerics(), cpoRegrResiduals(), cpoResponseFromSE(), cpoSample(), cpoScale(), cpoScaleMaxAbs(), cpoScaleRange(), cpoSelect(), cpoSmote(), cpoSpatialSign(), cpoTransformParams(), cpoWrap(), makeCPOCase(), makeCPOMultiplex()

Convert All Features to Numerics

Description

This is a CPOConstructor to be used to create a CPO. It is called like any R function and returns the created CPO.

Converts all feature columns to (integer) numeric columns by applying as.numeric to them.

Usage

cpoAsNumeric(
  id,
  export = "export.default",
  affect.type = NULL,
  affect.index = integer(0),
  affect.names = character(0),
  affect.pattern = NULL,
  affect.invert = FALSE,
  affect.pattern.ignore.case = FALSE,
  affect.pattern.perl = FALSE,
  affect.pattern.fixed = FALSE
)

Arguments

Value

[CPO].

General CPO info

This function creates a CPO object, which can be applied to Tasks, data.frames, link[mlr]{Learner}s and other CPO objects using the %>>% operator.

The parameters of this object can be changed after creation using the function setHyperPars. The other hyper-parameter manipulating functins, getHyperPars and getParamSet similarly work as one expects.

If the “id” parameter is given, the hyperparameters will have this id as aprefix; this will, however, not change the parameters of the creator function.

Calling a CPOConstructor

CPO constructor functions are called with optional values of parameters, and additional “special” optional values. The special optional values are the id parameter, and the affect.* parameters. The affect.* parameters enable the user to control which subset of a given dataset is affected. If no affect.* parameters are given, all data features are affected by default.

See Also

Other CPOs: cpoApplyFun(), cpoApplyFunRegrTarget(), cpoCache(), cpoCbind(), cpoCollapseFact(), cpoDropConstants(), cpoDropMostlyConstants(), cpoDummyEncode(), cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), cpoFixFactors(), cpoIca(), cpoImpactEncodeClassif(), cpoImpactEncodeRegr(), cpoImpute(), cpoImputeConstant(), cpoImputeHist(), cpoImputeLearner(), cpoImputeMax(), cpoImputeMean(), cpoImputeMedian(), cpoImputeMin(), cpoImputeMode(), cpoImputeNormal(), cpoImputeUniform(), cpoLogTrafoRegr(), cpoMakeCols(), cpoMissingIndicators(), cpoModelMatrix(), cpoOversample(), cpoPca(), cpoProbEncode(), cpoQuantileBinNumerics(), cpoRegrResiduals(), cpoResponseFromSE(), cpoSample(), cpoScale(), cpoScaleMaxAbs(), cpoScaleRange(), cpoSelect(), cpoSmote(), cpoSpatialSign(), cpoTransformParams(), cpoWrap(), makeCPOCase(), makeCPOMultiplex()

Caches the Result of CPO Transformations

Description

This is a CPOConstructor to be used to create a CPO. It is called like any R function and returns the created CPO.

Given a CPO to wrap, this caches an intermediate result (in fact, the retrafo object) whenever the CPO is applied to a Task or data.frame. This can reduce computation time when the same CPO is often applied to the same data, e.g. in a resampling or tuning evaluation.

The hyperparameters of the CPO are not exported, since in many cases changing the hyperparameters will also change the result and would defeat the point of caching. To switch between different settings of the same CPO, consider using cpoMultiplex.

The cache is kept in an environment; therefore, it does not communicate with other threads or processes when using parallelization at a level before the cache gets filled.

Caching needs the ‘digest’ package to be installed.

Usage

cpoCache(cpo = NULLCPO, cache.entries = 1024)

Arguments

Value

[CPO].

General CPO info

This function creates a CPO object, which can be applied to Tasks, data.frames, link[mlr]{Learner}s and other CPO objects using the %>>% operator.

The parameters of this object can be changed after creation using the function setHyperPars. The other hyper-parameter manipulating functins, getHyperPars and getParamSet similarly work as one expects.

If the “id” parameter is given, the hyperparameters will have this id as aprefix; this will, however, not change the parameters of the creator function.

Calling a CPOConstructor

CPO constructor functions are called with optional values of parameters, and additional “special” optional values. The special optional values are the id parameter, and the affect.* parameters. The affect.* parameters enable the user to control which subset of a given dataset is affected. If no affect.* parameters are given, all data features are affected by default.

See Also

Other CPOs: cpoApplyFun(), cpoApplyFunRegrTarget(), cpoAsNumeric(), cpoCbind(), cpoCollapseFact(), cpoDropConstants(), cpoDropMostlyConstants(), cpoDummyEncode(), cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), cpoFixFactors(), cpoIca(), cpoImpactEncodeClassif(), cpoImpactEncodeRegr(), cpoImpute(), cpoImputeConstant(), cpoImputeHist(), cpoImputeLearner(), cpoImputeMax(), cpoImputeMean(), cpoImputeMedian(), cpoImputeMin(), cpoImputeMode(), cpoImputeNormal(), cpoImputeUniform(), cpoLogTrafoRegr(), cpoMakeCols(), cpoMissingIndicators(), cpoModelMatrix(), cpoOversample(), cpoPca(), cpoProbEncode(), cpoQuantileBinNumerics(), cpoRegrResiduals(), cpoResponseFromSE(), cpoSample(), cpoScale(), cpoScaleMaxAbs(), cpoScaleRange(), cpoSelect(), cpoSmote(), cpoSpatialSign(), cpoTransformParams(), cpoWrap(), makeCPOCase(), makeCPOMultiplex()

“cbind” the Result of Multiple CPOs

Description

This is a CPOConstructor to be used to create a CPO. It is called like any R function and returns the created CPO.

Build a CPO that represents the operations of its input parameters, performed in parallel and put together column wise.

For example, to construct a Task that contains the original data, as well as the data after scaling, one could do

task %>>% cpoCbind(NULLCPO, cpoScale())

The result of cpoCbind is itself a CPO which exports its constituents' hyperparameters. CPOs with the same type / ID get combined automatically. To get networks, e.g. of the form

      ,-C--E-.
     /    /   \
A---B----D-----F---G

one coul use the code

initcpo = A %>>% B
route1 = initcpo %>>% D
route2 = cpoCbind(route1, initcpo %>>% C) %>>% E
result = cpoCbind(route1, route2) %>>% F %>>% G

cpoCbind finds common paths among its arguments and combines them into one operation. This saves computation and makes it possible for one exported hyperparameter to influence multiple of cpoCbind's inputs. However, if you want to use the same operation with different parameters on different parts of cpoCbind input, you must give these operations different IDs. If CPOs that could represent an identical CPO, with the same IDs (or both with IDs absent) but different parameter settings, affect.* settings or different parameter exportations occur, an error will be thrown.

Usage

cpoCbind(..., .cpos = list())

Arguments

Value

[CPO].

General CPO info

This function creates a CPO object, which can be applied to Tasks, data.frames, link[mlr]{Learner}s and other CPO objects using the %>>% operator.

The parameters of this object can be changed after creation using the function setHyperPars. The other hyper-parameter manipulating functins, getHyperPars and getParamSet similarly work as one expects.

If the “id” parameter is given, the hyperparameters will have this id as aprefix; this will, however, not change the parameters of the creator function.

Calling a CPOConstructor

CPO constructor functions are called with optional values of parameters, and additional “special” optional values. The special optional values are the id parameter, and the affect.* parameters. The affect.* parameters enable the user to control which subset of a given dataset is affected. If no affect.* parameters are given, all data features are affected by default.

See Also

Other CPOs: cpoApplyFun(), cpoApplyFunRegrTarget(), cpoAsNumeric(), cpoCache(), cpoCollapseFact(), cpoDropConstants(), cpoDropMostlyConstants(), cpoDummyEncode(), cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), cpoFixFactors(), cpoIca(), cpoImpactEncodeClassif(), cpoImpactEncodeRegr(), cpoImpute(), cpoImputeConstant(), cpoImputeHist(), cpoImputeLearner(), cpoImputeMax(), cpoImputeMean(), cpoImputeMedian(), cpoImputeMin(), cpoImputeMode(), cpoImputeNormal(), cpoImputeUniform(), cpoLogTrafoRegr(), cpoMakeCols(), cpoMissingIndicators(), cpoModelMatrix(), cpoOversample(), cpoPca(), cpoProbEncode(), cpoQuantileBinNumerics(), cpoRegrResiduals(), cpoResponseFromSE(), cpoSample(), cpoScale(), cpoScaleMaxAbs(), cpoScaleRange(), cpoSelect(), cpoSmote(), cpoSpatialSign(), cpoTransformParams(), cpoWrap(), makeCPOCase(), makeCPOMultiplex()

Other special CPOs: cpoTransformParams(), cpoWrap(), makeCPOCase(), makeCPOMultiplex()

Compine Rare Factors

Description

This is a CPOConstructor to be used to create a CPO. It is called like any R function and returns the created CPO.

Combine rare factor levels into a single factor level.

Usage

cpoCollapseFact(
  max.collapsed.class.prevalence = 0.1,
  id,
  export = "export.default",
  affect.type = NULL,
  affect.index = integer(0),
  affect.names = character(0),
  affect.pattern = NULL,
  affect.invert = FALSE,
  affect.pattern.ignore.case = FALSE,
  affect.pattern.perl = FALSE,
  affect.pattern.fixed = FALSE
)

Arguments

Value

[CPO].

General CPO info

This function creates a CPO object, which can be applied to Tasks, data.frames, link[mlr]{Learner}s and other CPO objects using the %>>% operator.

The parameters of this object can be changed after creation using the function setHyperPars. The other hyper-parameter manipulating functins, getHyperPars and getParamSet similarly work as one expects.

If the “id” parameter is given, the hyperparameters will have this id as aprefix; this will, however, not change the parameters of the creator function.

Calling a CPOConstructor

CPO constructor functions are called with optional values of parameters, and additional “special” optional values. The special optional values are the id parameter, and the affect.* parameters. The affect.* parameters enable the user to control which subset of a given dataset is affected. If no affect.* parameters are given, all data features are affected by default.

See Also

Other CPOs: cpoApplyFun(), cpoApplyFunRegrTarget(), cpoAsNumeric(), cpoCache(), cpoCbind(), cpoDropConstants(), cpoDropMostlyConstants(), cpoDummyEncode(), cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), cpoFixFactors(), cpoIca(), cpoImpactEncodeClassif(), cpoImpactEncodeRegr(), cpoImpute(), cpoImputeConstant(), cpoImputeHist(), cpoImputeLearner(), cpoImputeMax(), cpoImputeMean(), cpoImputeMedian(), cpoImputeMin(), cpoImputeMode(), cpoImputeNormal(), cpoImputeUniform(), cpoLogTrafoRegr(), cpoMakeCols(), cpoMissingIndicators(), cpoModelMatrix(), cpoOversample(), cpoPca(), cpoProbEncode(), cpoQuantileBinNumerics(), cpoRegrResiduals(), cpoResponseFromSE(), cpoSample(), cpoScale(), cpoScaleMaxAbs(), cpoScaleRange(), cpoSelect(), cpoSmote(), cpoSpatialSign(), cpoTransformParams(), cpoWrap(), makeCPOCase(), makeCPOMultiplex()

Drop Constant or Near-Constant Features

Description

This is a CPOConstructor to be used to create a CPO. It is called like any R function and returns the created CPO.

Drop all columns that are either constant, or close to constant for numerics, and columns that have only one value for factors or ordered columns.

Usage

cpoDropConstants(
  rel.tol = 1e-08,
  abs.tol = 1e-08,
  ignore.na = FALSE,
  id,
  export = "export.default",
  affect.type = NULL,
  affect.index = integer(0),
  affect.names = character(0),
  affect.pattern = NULL,
  affect.invert = FALSE,
  affect.pattern.ignore.case = FALSE,
  affect.pattern.perl = FALSE,
  affect.pattern.fixed = FALSE
)

Arguments

Value

[CPO].

General CPO info

This function creates a CPO object, which can be applied to Tasks, data.frames, link[mlr]{Learner}s and other CPO objects using the %>>% operator.

The parameters of this object can be changed after creation using the function setHyperPars. The other hyper-parameter manipulating functins, getHyperPars and getParamSet similarly work as one expects.

If the “id” parameter is given, the hyperparameters will have this id as aprefix; this will, however, not change the parameters of the creator function.

Calling a CPOConstructor

CPO constructor functions are called with optional values of parameters, and additional “special” optional values. The special optional values are the id parameter, and the affect.* parameters. The affect.* parameters enable the user to control which subset of a given dataset is affected. If no affect.* parameters are given, all data features are affected by default.

See Also

Other CPOs: cpoApplyFun(), cpoApplyFunRegrTarget(), cpoAsNumeric(), cpoCache(), cpoCbind(), cpoCollapseFact(), cpoDropMostlyConstants(), cpoDummyEncode(), cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), cpoFixFactors(), cpoIca(), cpoImpactEncodeClassif(), cpoImpactEncodeRegr(), cpoImpute(), cpoImputeConstant(), cpoImputeHist(), cpoImputeLearner(), cpoImputeMax(), cpoImputeMean(), cpoImputeMedian(), cpoImputeMin(), cpoImputeMode(), cpoImputeNormal(), cpoImputeUniform(), cpoLogTrafoRegr(), cpoMakeCols(), cpoMissingIndicators(), cpoModelMatrix(), cpoOversample(), cpoPca(), cpoProbEncode(), cpoQuantileBinNumerics(), cpoRegrResiduals(), cpoResponseFromSE(), cpoSample(), cpoScale(), cpoScaleMaxAbs(), cpoScaleRange(), cpoSelect(), cpoSmote(), cpoSpatialSign(), cpoTransformParams(), cpoWrap(), makeCPOCase(), makeCPOMultiplex()

Drop Constant or Near-Constant Features

Description

This is a CPOConstructor to be used to create a CPO. It is called like any R function and returns the created CPO.

Drop all columns that are mostly constant: Constant within tolerance with numerics, and and columns that have only one value for factors or ordered columns.

This CPO can also filter “mostly” constant Features: ones where at most a fraction of ratio samples differ from the mode value.

Usage

cpoDropMostlyConstants(
  ratio = 0,
  rel.tol = 1e-08,
  abs.tol = 1e-08,
  ignore.na = FALSE,
  id,
  export = "export.default",
  affect.type = NULL,
  affect.index = integer(0),
  affect.names = character(0),
  affect.pattern = NULL,
  affect.invert = FALSE,
  affect.pattern.ignore.case = FALSE,
  affect.pattern.perl = FALSE,
  affect.pattern.fixed = FALSE
)

Arguments

Value

[CPO].

General CPO info

This function creates a CPO object, which can be applied to Tasks, data.frames, link[mlr]{Learner}s and other CPO objects using the %>>% operator.

The parameters of this object can be changed after creation using the function setHyperPars. The other hyper-parameter manipulating functins, getHyperPars and getParamSet similarly work as one expects.

If the “id” parameter is given, the hyperparameters will have this id as aprefix; this will, however, not change the parameters of the creator function.

Calling a CPOConstructor

CPO constructor functions are called with optional values of parameters, and additional “special” optional values. The special optional values are the id parameter, and the affect.* parameters. The affect.* parameters enable the user to control which subset of a given dataset is affected. If no affect.* parameters are given, all data features are affected by default.

See Also

Other CPOs: cpoApplyFun(), cpoApplyFunRegrTarget(), cpoAsNumeric(), cpoCache(), cpoCbind(), cpoCollapseFact(), cpoDropConstants(), cpoDummyEncode(), cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), cpoFixFactors(), cpoIca(), cpoImpactEncodeClassif(), cpoImpactEncodeRegr(), cpoImpute(), cpoImputeConstant(), cpoImputeHist(), cpoImputeLearner(), cpoImputeMax(), cpoImputeMean(), cpoImputeMedian(), cpoImputeMin(), cpoImputeMode(), cpoImputeNormal(), cpoImputeUniform(), cpoLogTrafoRegr(), cpoMakeCols(), cpoMissingIndicators(), cpoModelMatrix(), cpoOversample(), cpoPca(), cpoProbEncode(), cpoQuantileBinNumerics(), cpoRegrResiduals(), cpoResponseFromSE(), cpoSample(), cpoScale(), cpoScaleMaxAbs(), cpoScaleRange(), cpoSelect(), cpoSmote(), cpoSpatialSign(), cpoTransformParams(), cpoWrap(), makeCPOCase(), makeCPOMultiplex()

CPO Dummy Encoder

Description

This is a CPOConstructor to be used to create a CPO. It is called like any R function and returns the created CPO.

Usage

cpoDummyEncode(
  reference.cat = FALSE,
  infixdot = FALSE,
  id,
  export = "export.default",
  affect.type = NULL,
  affect.index = integer(0),
  affect.names = character(0),
  affect.pattern = NULL,
  affect.invert = FALSE,
  affect.pattern.ignore.case = FALSE,
  affect.pattern.perl = FALSE,
  affect.pattern.fixed = FALSE
)

Arguments

Value

[CPO].

General CPO info

This function creates a CPO object, which can be applied to Tasks, data.frames, link[mlr]{Learner}s and other CPO objects using the %>>% operator.

The parameters of this object can be changed after creation using the function setHyperPars. The other hyper-parameter manipulating functins, getHyperPars and getParamSet similarly work as one expects.

If the “id” parameter is given, the hyperparameters will have this id as aprefix; this will, however, not change the parameters of the creator function.

Calling a CPOConstructor

CPO constructor functions are called with optional values of parameters, and additional “special” optional values. The special optional values are the id parameter, and the affect.* parameters. The affect.* parameters enable the user to control which subset of a given dataset is affected. If no affect.* parameters are given, all data features are affected by default.

See Also

Other CPOs: cpoApplyFun(), cpoApplyFunRegrTarget(), cpoAsNumeric(), cpoCache(), cpoCbind(), cpoCollapseFact(), cpoDropConstants(), cpoDropMostlyConstants(), cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), cpoFixFactors(), cpoIca(), cpoImpactEncodeClassif(), cpoImpactEncodeRegr(), cpoImpute(), cpoImputeConstant(), cpoImputeHist(), cpoImputeLearner(), cpoImputeMax(), cpoImputeMean(), cpoImputeMedian(), cpoImputeMin(), cpoImputeMode(), cpoImputeNormal(), cpoImputeUniform(), cpoLogTrafoRegr(), cpoMakeCols(), cpoMissingIndicators(), cpoModelMatrix(), cpoOversample(), cpoPca(), cpoProbEncode(), cpoQuantileBinNumerics(), cpoRegrResiduals(), cpoResponseFromSE(), cpoSample(), cpoScale(), cpoScaleMaxAbs(), cpoScaleRange(), cpoSelect(), cpoSmote(), cpoSpatialSign(), cpoTransformParams(), cpoWrap(), makeCPOCase(), makeCPOMultiplex()

Filter Features: “anova.test”

Description

This is a CPOConstructor to be used to create a CPO. It is called like any R function and returns the created CPO.

Filter “anova.test” is based on the Analysis of Variance (ANOVA) between feature and class. The value of the F-statistic is used as a measure of feature importance.

Usage

cpoFilterAnova(
  perc = NULL,
  abs = NULL,
  threshold = NULL,
  id,
  export = "export.default",
  affect.type = NULL,
  affect.index = integer(0),
  affect.names = character(0),
  affect.pattern = NULL,
  affect.invert = FALSE,
  affect.pattern.ignore.case = FALSE,
  affect.pattern.perl = FALSE,
  affect.pattern.fixed = FALSE
)

Arguments

Value

[CPO].

General CPO info

This function creates a CPO object, which can be applied to Tasks, data.frames, link[mlr]{Learner}s and other CPO objects using the %>>% operator.

The parameters of this object can be changed after creation using the function setHyperPars. The other hyper-parameter manipulating functins, getHyperPars and getParamSet similarly work as one expects.

If the “id” parameter is given, the hyperparameters will have this id as aprefix; this will, however, not change the parameters of the creator function.

Calling a CPOConstructor

CPO constructor functions are called with optional values of parameters, and additional “special” optional values. The special optional values are the id parameter, and the affect.* parameters. The affect.* parameters enable the user to control which subset of a given dataset is affected. If no affect.* parameters are given, all data features are affected by default.

See Also

Other filter: cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), randomForestSRC_filters

Other CPOs: cpoApplyFun(), cpoApplyFunRegrTarget(), cpoAsNumeric(), cpoCache(), cpoCbind(), cpoCollapseFact(), cpoDropConstants(), cpoDropMostlyConstants(), cpoDummyEncode(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), cpoFixFactors(), cpoIca(), cpoImpactEncodeClassif(), cpoImpactEncodeRegr(), cpoImpute(), cpoImputeConstant(), cpoImputeHist(), cpoImputeLearner(), cpoImputeMax(), cpoImputeMean(), cpoImputeMedian(), cpoImputeMin(), cpoImputeMode(), cpoImputeNormal(), cpoImputeUniform(), cpoLogTrafoRegr(), cpoMakeCols(), cpoMissingIndicators(), cpoModelMatrix(), cpoOversample(), cpoPca(), cpoProbEncode(), cpoQuantileBinNumerics(), cpoRegrResiduals(), cpoResponseFromSE(), cpoSample(), cpoScale(), cpoScaleMaxAbs(), cpoScaleRange(), cpoSelect(), cpoSmote(), cpoSpatialSign(), cpoTransformParams(), cpoWrap(), makeCPOCase(), makeCPOMultiplex()

Filter Features: “carscore”

Description

This is a CPOConstructor to be used to create a CPO. It is called like any R function and returns the created CPO.

Filter “carscore” determines the “Correlation-Adjusted (marginal) coRelation scores” (short CAR scores). The CAR scores for a set of features are defined as the correlations between the target and the decorrelated features.

Usage

cpoFilterCarscore(
  diagonal = FALSE,
  perc = NULL,
  abs = NULL,
  threshold = NULL,
  id,
  export = "export.default",
  affect.type = NULL,
  affect.index = integer(0),
  affect.names = character(0),
  affect.pattern = NULL,
  affect.invert = FALSE,
  affect.pattern.ignore.case = FALSE,
  affect.pattern.perl = FALSE,
  affect.pattern.fixed = FALSE
)

Arguments

Value

[CPO].

General CPO info

This function creates a CPO object, which can be applied to Tasks, data.frames, link[mlr]{Learner}s and other CPO objects using the %>>% operator.

The parameters of this object can be changed after creation using the function setHyperPars. The other hyper-parameter manipulating functins, getHyperPars and getParamSet similarly work as one expects.

If the “id” parameter is given, the hyperparameters will have this id as aprefix; this will, however, not change the parameters of the creator function.

Calling a CPOConstructor

CPO constructor functions are called with optional values of parameters, and additional “special” optional values. The special optional values are the id parameter, and the affect.* parameters. The affect.* parameters enable the user to control which subset of a given dataset is affected. If no affect.* parameters are given, all data features are affected by default.

See Also

Other filter: cpoFilterAnova(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), randomForestSRC_filters

Other CPOs: cpoApplyFun(), cpoApplyFunRegrTarget(), cpoAsNumeric(), cpoCache(), cpoCbind(), cpoCollapseFact(), cpoDropConstants(), cpoDropMostlyConstants(), cpoDummyEncode(), cpoFilterAnova(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), cpoFixFactors(), cpoIca(), cpoImpactEncodeClassif(), cpoImpactEncodeRegr(), cpoImpute(), cpoImputeConstant(), cpoImputeHist(), cpoImputeLearner(), cpoImputeMax(), cpoImputeMean(), cpoImputeMedian(), cpoImputeMin(), cpoImputeMode(), cpoImputeNormal(), cpoImputeUniform(), cpoLogTrafoRegr(), cpoMakeCols(), cpoMissingIndicators(), cpoModelMatrix(), cpoOversample(), cpoPca(), cpoProbEncode(), cpoQuantileBinNumerics(), cpoRegrResiduals(), cpoResponseFromSE(), cpoSample(), cpoScale(), cpoScaleMaxAbs(), cpoScaleRange(), cpoSelect(), cpoSmote(), cpoSpatialSign(), cpoTransformParams(), cpoWrap(), makeCPOCase(), makeCPOMultiplex()

Filter Features: “chi.squared”

Description

This is a CPOConstructor to be used to create a CPO. It is called like any R function and returns the created CPO.

The chi-square test is a statistical test of independence to determine whether two variables are independent. Filter “chi.squared” applies this test in the following way. For each feature the chi-square test statistic is computed checking if there is a dependency between the feature and the target variable. Low values of the test statistic indicate a poor relationship. High values, i.e., high dependency identifies a feature as more important.

Usage

cpoFilterChiSquared(
  perc = NULL,
  abs = NULL,
  threshold = NULL,
  id,
  export = "export.default",
  affect.type = NULL,
  affect.index = integer(0),
  affect.names = character(0),
  affect.pattern = NULL,
  affect.invert = FALSE,
  affect.pattern.ignore.case = FALSE,
  affect.pattern.perl = FALSE,
  affect.pattern.fixed = FALSE
)

Arguments

Value

[CPO].

General CPO info

This function creates a CPO object, which can be applied to Tasks, data.frames, link[mlr]{Learner}s and other CPO objects using the %>>% operator.

The parameters of this object can be changed after creation using the function setHyperPars. The other hyper-parameter manipulating functins, getHyperPars and getParamSet similarly work as one expects.

If the “id” parameter is given, the hyperparameters will have this id as aprefix; this will, however, not change the parameters of the creator function.

Calling a CPOConstructor

CPO constructor functions are called with optional values of parameters, and additional “special” optional values. The special optional values are the id parameter, and the affect.* parameters. The affect.* parameters enable the user to control which subset of a given dataset is affected. If no affect.* parameters are given, all data features are affected by default.

See Also

Other filter: cpoFilterAnova(), cpoFilterCarscore(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), randomForestSRC_filters

Other CPOs: cpoApplyFun(), cpoApplyFunRegrTarget(), cpoAsNumeric(), cpoCache(), cpoCbind(), cpoCollapseFact(), cpoDropConstants(), cpoDropMostlyConstants(), cpoDummyEncode(), cpoFilterAnova(), cpoFilterCarscore(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), cpoFixFactors(), cpoIca(), cpoImpactEncodeClassif(), cpoImpactEncodeRegr(), cpoImpute(), cpoImputeConstant(), cpoImputeHist(), cpoImputeLearner(), cpoImputeMax(), cpoImputeMean(), cpoImputeMedian(), cpoImputeMin(), cpoImputeMode(), cpoImputeNormal(), cpoImputeUniform(), cpoLogTrafoRegr(), cpoMakeCols(), cpoMissingIndicators(), cpoModelMatrix(), cpoOversample(), cpoPca(), cpoProbEncode(), cpoQuantileBinNumerics(), cpoRegrResiduals(), cpoResponseFromSE(), cpoSample(), cpoScale(), cpoScaleMaxAbs(), cpoScaleRange(), cpoSelect(), cpoSmote(), cpoSpatialSign(), cpoTransformParams(), cpoWrap(), makeCPOCase(), makeCPOMultiplex()

Filter Features by Thresholding Filter Values

Description

This is a CPOConstructor to be used to create a CPO. It is called like any R function and returns the created CPO.

First, calls generateFilterValuesData. Features are then selected via select and val.

Usage

cpoFilterFeatures(
  method = "randomForestSRC.rfsrc",
  fval = NULL,
  perc = NULL,
  abs = NULL,
  threshold = NULL,
  filter.args = list(),
  id,
  export = "export.default",
  affect.type = NULL,
  affect.index = integer(0),
  affect.names = character(0),
  affect.pattern = NULL,
  affect.invert = FALSE,
  affect.pattern.ignore.case = FALSE,
  affect.pattern.perl = FALSE,
  affect.pattern.fixed = FALSE
)

Arguments

Value

[CPO].

General CPO info

This function creates a CPO object, which can be applied to Tasks, data.frames, link[mlr]{Learner}s and other CPO objects using the %>>% operator.

The parameters of this object can be changed after creation using the function setHyperPars. The other hyper-parameter manipulating functins, getHyperPars and getParamSet similarly work as one expects.

If the “id” parameter is given, the hyperparameters will have this id as aprefix; this will, however, not change the parameters of the creator function.

Calling a CPOConstructor

CPO constructor functions are called with optional values of parameters, and additional “special” optional values. The special optional values are the id parameter, and the affect.* parameters. The affect.* parameters enable the user to control which subset of a given dataset is affected. If no affect.* parameters are given, all data features are affected by default.

See Also

Other filter: cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), randomForestSRC_filters

Other CPOs: cpoApplyFun(), cpoApplyFunRegrTarget(), cpoAsNumeric(), cpoCache(), cpoCbind(), cpoCollapseFact(), cpoDropConstants(), cpoDropMostlyConstants(), cpoDummyEncode(), cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), cpoFixFactors(), cpoIca(), cpoImpactEncodeClassif(), cpoImpactEncodeRegr(), cpoImpute(), cpoImputeConstant(), cpoImputeHist(), cpoImputeLearner(), cpoImputeMax(), cpoImputeMean(), cpoImputeMedian(), cpoImputeMin(), cpoImputeMode(), cpoImputeNormal(), cpoImputeUniform(), cpoLogTrafoRegr(), cpoMakeCols(), cpoMissingIndicators(), cpoModelMatrix(), cpoOversample(), cpoPca(), cpoProbEncode(), cpoQuantileBinNumerics(), cpoRegrResiduals(), cpoResponseFromSE(), cpoSample(), cpoScale(), cpoScaleMaxAbs(), cpoScaleRange(), cpoSelect(), cpoSmote(), cpoSpatialSign(), cpoTransformParams(), cpoWrap(), makeCPOCase(), makeCPOMultiplex()

Other filter: cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), randomForestSRC_filters

Filter Features: “gain.ratio”

Description

This is a CPOConstructor to be used to create a CPO. It is called like any R function and returns the created CPO.

Filter “gain.ratio” uses the entropy-based information gain ratio between each feature and target individually as an importance measure.

Usage

cpoFilterGainRatio(
  perc = NULL,
  abs = NULL,
  threshold = NULL,
  id,
  export = "export.default",
  affect.type = NULL,
  affect.index = integer(0),
  affect.names = character(0),
  affect.pattern = NULL,
  affect.invert = FALSE,
  affect.pattern.ignore.case = FALSE,
  affect.pattern.perl = FALSE,
  affect.pattern.fixed = FALSE
)

Arguments

Value

[CPO].

General CPO info

This function creates a CPO object, which can be applied to Tasks, data.frames, link[mlr]{Learner}s and other CPO objects using the %>>% operator.

The parameters of this object can be changed after creation using the function setHyperPars. The other hyper-parameter manipulating functins, getHyperPars and getParamSet similarly work as one expects.

If the “id” parameter is given, the hyperparameters will have this id as aprefix; this will, however, not change the parameters of the creator function.

Calling a CPOConstructor

CPO constructor functions are called with optional values of parameters, and additional “special” optional values. The special optional values are the id parameter, and the affect.* parameters. The affect.* parameters enable the user to control which subset of a given dataset is affected. If no affect.* parameters are given, all data features are affected by default.

See Also

Other filter: cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), randomForestSRC_filters

Other CPOs: cpoApplyFun(), cpoApplyFunRegrTarget(), cpoAsNumeric(), cpoCache(), cpoCbind(), cpoCollapseFact(), cpoDropConstants(), cpoDropMostlyConstants(), cpoDummyEncode(), cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), cpoFixFactors(), cpoIca(), cpoImpactEncodeClassif(), cpoImpactEncodeRegr(), cpoImpute(), cpoImputeConstant(), cpoImputeHist(), cpoImputeLearner(), cpoImputeMax(), cpoImputeMean(), cpoImputeMedian(), cpoImputeMin(), cpoImputeMode(), cpoImputeNormal(), cpoImputeUniform(), cpoLogTrafoRegr(), cpoMakeCols(), cpoMissingIndicators(), cpoModelMatrix(), cpoOversample(), cpoPca(), cpoProbEncode(), cpoQuantileBinNumerics(), cpoRegrResiduals(), cpoResponseFromSE(), cpoSample(), cpoScale(), cpoScaleMaxAbs(), cpoScaleRange(), cpoSelect(), cpoSmote(), cpoSpatialSign(), cpoTransformParams(), cpoWrap(), makeCPOCase(), makeCPOMultiplex()

Filter Features: “information.gain”

Description

This is a CPOConstructor to be used to create a CPO. It is called like any R function and returns the created CPO.

Filter “information.gain” uses the entropy-based information gain between each feature and target individually as an importance measure.

Usage

cpoFilterInformationGain(
  perc = NULL,
  abs = NULL,
  threshold = NULL,
  id,
  export = "export.default",
  affect.type = NULL,
  affect.index = integer(0),
  affect.names = character(0),
  affect.pattern = NULL,
  affect.invert = FALSE,
  affect.pattern.ignore.case = FALSE,
  affect.pattern.perl = FALSE,
  affect.pattern.fixed = FALSE
)

Arguments

Value

[CPO].

General CPO info

This function creates a CPO object, which can be applied to Tasks, data.frames, link[mlr]{Learner}s and other CPO objects using the %>>% operator.

The parameters of this object can be changed after creation using the function setHyperPars. The other hyper-parameter manipulating functins, getHyperPars and getParamSet similarly work as one expects.

If the “id” parameter is given, the hyperparameters will have this id as aprefix; this will, however, not change the parameters of the creator function.

Calling a CPOConstructor

CPO constructor functions are called with optional values of parameters, and additional “special” optional values. The special optional values are the id parameter, and the affect.* parameters. The affect.* parameters enable the user to control which subset of a given dataset is affected. If no affect.* parameters are given, all data features are affected by default.

See Also

Other filter: cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), randomForestSRC_filters

Other CPOs: cpoApplyFun(), cpoApplyFunRegrTarget(), cpoAsNumeric(), cpoCache(), cpoCbind(), cpoCollapseFact(), cpoDropConstants(), cpoDropMostlyConstants(), cpoDummyEncode(), cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), cpoFixFactors(), cpoIca(), cpoImpactEncodeClassif(), cpoImpactEncodeRegr(), cpoImpute(), cpoImputeConstant(), cpoImputeHist(), cpoImputeLearner(), cpoImputeMax(), cpoImputeMean(), cpoImputeMedian(), cpoImputeMin(), cpoImputeMode(), cpoImputeNormal(), cpoImputeUniform(), cpoLogTrafoRegr(), cpoMakeCols(), cpoMissingIndicators(), cpoModelMatrix(), cpoOversample(), cpoPca(), cpoProbEncode(), cpoQuantileBinNumerics(), cpoRegrResiduals(), cpoResponseFromSE(), cpoSample(), cpoScale(), cpoScaleMaxAbs(), cpoScaleRange(), cpoSelect(), cpoSmote(), cpoSpatialSign(), cpoTransformParams(), cpoWrap(), makeCPOCase(), makeCPOMultiplex()

Filter Features: “kruskal.test”

Description

This is a CPOConstructor to be used to create a CPO. It is called like any R function and returns the created CPO.

Filter “kruskal.test” applies a Kruskal-Wallis rank sum test of the null hypothesis that the location parameters of the distribution of a feature are the same in each class and considers the test statistic as an variable importance measure: if the location parameters do not differ in at least one case, i.e., the null hypothesis cannot be rejected, there is little evidence that the corresponding feature is suitable for classification.

Usage

cpoFilterKruskal(
  perc = NULL,
  abs = NULL,
  threshold = NULL,
  id,
  export = "export.default",
  affect.type = NULL,
  affect.index = integer(0),
  affect.names = character(0),
  affect.pattern = NULL,
  affect.invert = FALSE,
  affect.pattern.ignore.case = FALSE,
  affect.pattern.perl = FALSE,
  affect.pattern.fixed = FALSE
)

Arguments

Value

[CPO].

General CPO info

This function creates a CPO object, which can be applied to Tasks, data.frames, link[mlr]{Learner}s and other CPO objects using the %>>% operator.

The parameters of this object can be changed after creation using the function setHyperPars. The other hyper-parameter manipulating functins, getHyperPars and getParamSet similarly work as one expects.

If the “id” parameter is given, the hyperparameters will have this id as aprefix; this will, however, not change the parameters of the creator function.

Calling a CPOConstructor

CPO constructor functions are called with optional values of parameters, and additional “special” optional values. The special optional values are the id parameter, and the affect.* parameters. The affect.* parameters enable the user to control which subset of a given dataset is affected. If no affect.* parameters are given, all data features are affected by default.

See Also

Other filter: cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), randomForestSRC_filters

Other CPOs: cpoApplyFun(), cpoApplyFunRegrTarget(), cpoAsNumeric(), cpoCache(), cpoCbind(), cpoCollapseFact(), cpoDropConstants(), cpoDropMostlyConstants(), cpoDummyEncode(), cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), cpoFixFactors(), cpoIca(), cpoImpactEncodeClassif(), cpoImpactEncodeRegr(), cpoImpute(), cpoImputeConstant(), cpoImputeHist(), cpoImputeLearner(), cpoImputeMax(), cpoImputeMean(), cpoImputeMedian(), cpoImputeMin(), cpoImputeMode(), cpoImputeNormal(), cpoImputeUniform(), cpoLogTrafoRegr(), cpoMakeCols(), cpoMissingIndicators(), cpoModelMatrix(), cpoOversample(), cpoPca(), cpoProbEncode(), cpoQuantileBinNumerics(), cpoRegrResiduals(), cpoResponseFromSE(), cpoSample(), cpoScale(), cpoScaleMaxAbs(), cpoScaleRange(), cpoSelect(), cpoSmote(), cpoSpatialSign(), cpoTransformParams(), cpoWrap(), makeCPOCase(), makeCPOMultiplex()

Filter Features: “linear.correlation”

Description

This is a CPOConstructor to be used to create a CPO. It is called like any R function and returns the created CPO.

The Pearson correlation between each feature and the target is used as an indicator of feature importance. Rows with NA values are not taken into consideration.

Usage

cpoFilterLinearCorrelation(
  perc = NULL,
  abs = NULL,
  threshold = NULL,
  id,
  export = "export.default",
  affect.type = NULL,
  affect.index = integer(0),
  affect.names = character(0),
  affect.pattern = NULL,
  affect.invert = FALSE,
  affect.pattern.ignore.case = FALSE,
  affect.pattern.perl = FALSE,
  affect.pattern.fixed = FALSE
)

Arguments

Value

[CPO].

General CPO info

This function creates a CPO object, which can be applied to Tasks, data.frames, link[mlr]{Learner}s and other CPO objects using the %>>% operator.

The parameters of this object can be changed after creation using the function setHyperPars. The other hyper-parameter manipulating functins, getHyperPars and getParamSet similarly work as one expects.

If the “id” parameter is given, the hyperparameters will have this id as aprefix; this will, however, not change the parameters of the creator function.

Calling a CPOConstructor

CPO constructor functions are called with optional values of parameters, and additional “special” optional values. The special optional values are the id parameter, and the affect.* parameters. The affect.* parameters enable the user to control which subset of a given dataset is affected. If no affect.* parameters are given, all data features are affected by default.

See Also

Other filter: cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), randomForestSRC_filters

Other CPOs: cpoApplyFun(), cpoApplyFunRegrTarget(), cpoAsNumeric(), cpoCache(), cpoCbind(), cpoCollapseFact(), cpoDropConstants(), cpoDropMostlyConstants(), cpoDummyEncode(), cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterMrmr(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), cpoFixFactors(), cpoIca(), cpoImpactEncodeClassif(), cpoImpactEncodeRegr(), cpoImpute(), cpoImputeConstant(), cpoImputeHist(), cpoImputeLearner(), cpoImputeMax(), cpoImputeMean(), cpoImputeMedian(), cpoImputeMin(), cpoImputeMode(), cpoImputeNormal(), cpoImputeUniform(), cpoLogTrafoRegr(), cpoMakeCols(), cpoMissingIndicators(), cpoModelMatrix(), cpoOversample(), cpoPca(), cpoProbEncode(), cpoQuantileBinNumerics(), cpoRegrResiduals(), cpoResponseFromSE(), cpoSample(), cpoScale(), cpoScaleMaxAbs(), cpoScaleRange(), cpoSelect(), cpoSmote(), cpoSpatialSign(), cpoTransformParams(), cpoWrap(), makeCPOCase(), makeCPOMultiplex()

Filter Features: “mrmr”

Description

This is a CPOConstructor to be used to create a CPO. It is called like any R function and returns the created CPO.

Minimum redundancy, maximum relevance filter “mrmr” computes the mutual information between the target and each individual feature minus the average mutual information of previously selected features and this feature using the mRMRe package.

Usage

cpoFilterMrmr(
  perc = NULL,
  abs = NULL,
  threshold = NULL,
  id,
  export = "export.default",
  affect.type = NULL,
  affect.index = integer(0),
  affect.names = character(0),
  affect.pattern = NULL,
  affect.invert = FALSE,
  affect.pattern.ignore.case = FALSE,
  affect.pattern.perl = FALSE,
  affect.pattern.fixed = FALSE
)

Arguments

Value

[CPO].

General CPO info

This function creates a CPO object, which can be applied to Tasks, data.frames, link[mlr]{Learner}s and other CPO objects using the %>>% operator.

The parameters of this object can be changed after creation using the function setHyperPars. The other hyper-parameter manipulating functins, getHyperPars and getParamSet similarly work as one expects.

If the “id” parameter is given, the hyperparameters will have this id as aprefix; this will, however, not change the parameters of the creator function.

Calling a CPOConstructor

CPO constructor functions are called with optional values of parameters, and additional “special” optional values. The special optional values are the id parameter, and the affect.* parameters. The affect.* parameters enable the user to control which subset of a given dataset is affected. If no affect.* parameters are given, all data features are affected by default.

See Also

Other filter: cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), randomForestSRC_filters

Other CPOs: cpoApplyFun(), cpoApplyFunRegrTarget(), cpoAsNumeric(), cpoCache(), cpoCbind(), cpoCollapseFact(), cpoDropConstants(), cpoDropMostlyConstants(), cpoDummyEncode(), cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterOneR(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), cpoFixFactors(), cpoIca(), cpoImpactEncodeClassif(), cpoImpactEncodeRegr(), cpoImpute(), cpoImputeConstant(), cpoImputeHist(), cpoImputeLearner(), cpoImputeMax(), cpoImputeMean(), cpoImputeMedian(), cpoImputeMin(), cpoImputeMode(), cpoImputeNormal(), cpoImputeUniform(), cpoLogTrafoRegr(), cpoMakeCols(), cpoMissingIndicators(), cpoModelMatrix(), cpoOversample(), cpoPca(), cpoProbEncode(), cpoQuantileBinNumerics(), cpoRegrResiduals(), cpoResponseFromSE(), cpoSample(), cpoScale(), cpoScaleMaxAbs(), cpoScaleRange(), cpoSelect(), cpoSmote(), cpoSpatialSign(), cpoTransformParams(), cpoWrap(), makeCPOCase(), makeCPOMultiplex()

Filter Features: “oneR”

Description

This is a CPOConstructor to be used to create a CPO. It is called like any R function and returns the created CPO.

Filter “oneR” makes use of a simple “One-Rule” (OneR) learner to determine feature importance. For this purpose the OneR learner generates one simple association rule for each feature in the data individually and computes the total error. The lower the error value the more important the correspoding feature.

Usage

cpoFilterOneR(
  perc = NULL,
  abs = NULL,
  threshold = NULL,
  id,
  export = "export.default",
  affect.type = NULL,
  affect.index = integer(0),
  affect.names = character(0),
  affect.pattern = NULL,
  affect.invert = FALSE,
  affect.pattern.ignore.case = FALSE,
  affect.pattern.perl = FALSE,
  affect.pattern.fixed = FALSE
)

Arguments

Value

[CPO].

General CPO info

This function creates a CPO object, which can be applied to Tasks, data.frames, link[mlr]{Learner}s and other CPO objects using the %>>% operator.

The parameters of this object can be changed after creation using the function setHyperPars. The other hyper-parameter manipulating functins, getHyperPars and getParamSet similarly work as one expects.

If the “id” parameter is given, the hyperparameters will have this id as aprefix; this will, however, not change the parameters of the creator function.

Calling a CPOConstructor

CPO constructor functions are called with optional values of parameters, and additional “special” optional values. The special optional values are the id parameter, and the affect.* parameters. The affect.* parameters enable the user to control which subset of a given dataset is affected. If no affect.* parameters are given, all data features are affected by default.

See Also

Other filter: cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), randomForestSRC_filters

Other CPOs: cpoApplyFun(), cpoApplyFunRegrTarget(), cpoAsNumeric(), cpoCache(), cpoCbind(), cpoCollapseFact(), cpoDropConstants(), cpoDropMostlyConstants(), cpoDummyEncode(), cpoFilterAnova(), cpoFilterCarscore(), cpoFilterChiSquared(), cpoFilterFeatures(), cpoFilterGainRatio(), cpoFilterInformationGain(), cpoFilterKruskal(), cpoFilterLinearCorrelation(), cpoFilterMrmr(), cpoFilterPermutationImportance(), cpoFilterRankCorrelation(), cpoFilterRelief(), cpoFilterRfCImportance(), cpoFilterRfImportance(), cpoFilterRfSRCImportance(), cpoFilterSymmetricalUncertainty(), cpoFilterUnivariate(), cpoFilterVariance(), cpoFixFactors(), cpoIca(), cpoImpactEncodeClassif(), cpoImpactEncodeRegr(), cpoImpute(), cpoImputeConstant(), cpoImputeHist(), cpoImputeLearner(), cpoImputeMax(), cpoImputeMean(), cpoImputeMedian(), cpoImputeMin(), cpoImputeMode(), cpoImputeNormal(), cpoImputeUniform(), cpoLogTrafoRegr(), cpoMakeCols(), cpoMissingIndicators(), cpoModelMatrix(), cpoOversample(), cpoPca(), cpoProbEncode(), cpoQuantileBinNumerics(), cpoRegrResiduals(), cpoResponseFromSE(), cpoSample(), cpoScale(), cpoScaleMaxAbs(), cpoScaleRange(), cpoSelect(), cpoSmote(), cpoSpatialSign(), cpoTransformParams(), cpoWrap(), makeCPOCase(), makeCPOMultiplex()

Filter Features: “permutation.importance”

Description

This is a CPOConstructor to be used to create a CPO. It is called like any R function and returns the created CPO.

Filter “permutation.importance” computes a loss function between predictions made by a learner before and after a feature is permuted.

Usage

cpoFilterPermutationImportance(
  perc = NULL,
  abs = NULL,
  threshold = NULL,
  imp.learner,
  contrast = function(x, y) {
     x - y
 },
  measure = NULL,
  aggregation = function(x, ...) UseMethod("mean"),
  nmc = 50,
  replace = FALSE,
  id,
  export = "export.default",
  affect.type = NULL,
  affect.index = integer(0),
  affect.names = character(0),
  affect.pattern = NULL,
  affect.invert = FALSE,
  affect.pattern.ignore.case = FALSE,
  affect.pattern.perl = FALSE,
  affect.pattern.fixed = FALSE
)

Arguments