Creates a Subgroup Discovery Task

Description

Performs subgroup discovery according to the given task.

Usage

CreateSDTask(source, target, config = SDTaskConfig())

Arguments

See Also

DiscoverSubgroups. DiscoverSubgroupsByTask SDTaskConfig

Examples

# creating a task
data(credit.data)

# task with binary target
task <- CreateSDTask(credit.data, as.target("class", "good"))

# task with numeric target
taskNum <- CreateSDTask(credit.data, as.target("credit_amount"))

Performs Subgroup Discovery

Description

Performs subgroup discovery according to the given target and the configuration on the data.

Usage

DiscoverSubgroups(source, target, config= SDTaskConfig(), as.df=FALSE)

Arguments

See Also

DiscoverSubgroupsByTask. as.target CreateSDTask SDTaskConfig

Examples

# subgroup discovery on a data.frame, for binary target
data(credit.data)
result1 <- DiscoverSubgroups(
    credit.data, as.target("class", "good"), new("SDTaskConfig",
    attributes=c("checking_status", "credit_amount", "employment", "purpose")))
result2 <- DiscoverSubgroups(
    credit.data, as.target("class", "good"), new("SDTaskConfig",
    attributes=c("checking_status", "employment")))

ToDataFrame(result1)
ToDataFrame(result2)

# subgroup discovery for numeric target variable
result3 <- DiscoverSubgroups(
    credit.data, as.target("credit_amount"), new("SDTaskConfig",
    attributes=c("checking_status", "employment")))

ToDataFrame(result3)

Performs Subgroup Discovery for a given Task

Description

Performs subgroup discovery according to the given task.

Usage

DiscoverSubgroupsByTask(task, as.df=FALSE)

Arguments

See Also

DiscoverSubgroups. CreateSDTask

Examples

# creating a task
data(credit.data)
task <- CreateSDTask(
    credit.data, as.target("class", "bad"), SDTaskConfig(
    attributes=c("checking_status", "employment"))) 
taskNum <- CreateSDTask(
    credit.data, as.target("credit_amount"), SDTaskConfig(
    attributes=c("checking_status", "employment")))

# running the tasks
DiscoverSubgroupsByTask(task)
DiscoverSubgroupsByTask(taskNum)

Class “Pattern” — A Simple Subgroup Description Container

Description

A Simple Container holding the results (subgroups, description and parameters) for the Subgroup and Pattern Mining Algorithms

Objects from the Class

Objects are created by calls of the form new("Pattern", ...).

Slots

description:

The subgroup description, as a character vector.

selectors:

The subgroup description, given as a list of (simple) selection expressions, where the 'key' is the attribute and the 'value' is the value.

quality:

The numeric value denoting the quality of the subgroup pattern as determined by the applied quality function.

size:

The size of the subgroup.

parameters

Additional quality parameters of the subgroup.

See Also

DiscoverSubgroups. DiscoverSubgroupsByTask CreateSDTask

Creates a Subgroup Discovery Task Configuration

Description

Creates a subgroup discovery task configuration, that is, an instance of SDTaskConfig.

Class “SDTaskConfig” — A Set of Configuration Settings

Description

A Set of Configuration Settings for the Subgroup and Pattern Mining Algorithms

Objects from the Class

Objects are created by calls of the form SDTaskConfig(...).

Slots

attributes:

The list of attributes to consider for mining. Either a vector of attribute names, or NULL (the default), which includes all attributes.

discretize:

Boolean, indicating whether to (automatically) discretize numeric attributes (default discretize=TRUE. Depends on parameter nbins. Either creates distinct values, if their number in the dataset is <= nbins, or applies equal-frequency discretization for the respective numeric attribute.

method:

A mining method; one of Beam-Search beam, BSD bsd, SD-Map sdmap, SD-Map enabling internal disjunctions sdmap-dis. The default is method = "sdmap".

nbins:

Specifies the number of bins to be used when discretizing numeric attributes (see discretize above).

qf:

A quality function; one of: Adjusted Residuals ares, Binomial Test bin, Chi-Square Test chi2, Gain gain, Lift lift, Piatetsky-Shapiro ps, Relative Gain relgain, Weighted Relative Accuracy wracc. The default is qf = "ps".

k:

The maximum number (top-k) of patterns to discover, i.e., the best k rules according to the selected quality function. The default is k = 20

minqual:

The minimal quality (default minqual = 0).

minsize:

The minimal size of a subgroup (as an integer) (minimal coverage of database records, default minsize = 0).

mintp:

The minimal true positive (tp) threshold, an integer (minimal (absolute) number of true positives in a subgroup, relevant for binary target concepts only), defaults to mintp = 0

.

maxlen:

The maximal length of a description of a pattern, i.e., the maximal number of conjunctions. This impacts both understandability and efficiency. Simpler rules are easier to understand, and a small maxlen will restrict the search space (default maxlen = 7).

nodefaults:

Ignore default values, i.e., do not include the respective first value (with index 0) of each attribute (default nodefaults=FALSE, i.e., include all values).

relfilter:

Controls, whether irrelevant patterns are filtered during pattern mining; negatively impacts performance (default relfilter = FALSE)).

postfilter:

Controls, whether a post-processing filter is applied; one (or a vector) of: Minimum Improvement (Global) min-improve-global, checks the patterns against all possible generalizations, Minimum Improvement (Pattern Set) min-improve-set, checks the patterns against all their generalizations in the result set, Relevancy Filter relevancy, removes patterns that are strictly irrelevant, Significant Improvement (Global) sig-improve-global, removes patterns that do not significantly improve (default 0.01 level) w.r.t. all their possible generalizations, Significant Improvement (Set) sig-improve-set, removes patterns that do not significantly improve (default 0.01 level) w.r.t. all generalizations in the result set, Weighted Covering weighted-covering, performs weighted covering on the data in order to select a covering set of subgroups while reducing the overlap on the data. By default no postfilter is set, i.e., postfilter = "".

parfilter:

Provides the minimal improvement value for the postfilter (for min-improve-* filters), or the significance level (P) for sig-improve-* filters.

See Also

DiscoverSubgroups. DiscoverSubgroupsByTask CreateSDTask

Transforms patterns into a data frame

Description

Transforms a list/vector of patterns into a data frame for inspection and analysis.

Usage

ToDataFrame(patterns, ndigits = 2)

Arguments

See Also

DiscoverSubgroups.

Constructs a target variable (for subgroup discovery)

Description

Constructs a target variable, i.e., an object suitable to be passed to DiscoverSubgroups or CreateSDTask.

Usage

as.target(attribute, value=NULL)

Arguments

See Also

DiscoverSubgroups.

Examples

# creating a target variable
# binary:
as.target("class", "true")

#numeric:
as.target("numeric_class")

Statlog (German Credit Data) Data Set

Description

This dataset classifies people described by a set of attributes as good or bad credit risks.

Usage

data(credit.data)

Format

A vector containing 1000 observations.

Source

UCI Repository, https://archive.ics.uci.edu/ml/datasets/Statlog+(German+Credit+Data).

Tests whether a pattern and a data list (row of a data frame) match

Description

Tests whether a pattern and a data list (row of a data frame) match, e.g., for implementing classification methods.

Usage

is.pattern.matching(pattern, data.list)

Arguments

See Also

Pattern-class.

rsubgroup Package - Algorithms and Tools for Efficient Subgroup Discovery and Analytics

Description

The rsubgroup package contains a set of efficient and effective tools and algorithms for subgroup discovery and analytics. The package integrates an R interface to the org.vikamine.kernel library of the VIKAMINE system (http://www.vikamine.org).

Note: rsubgroup uses rJava. To set the maximum available heap space for Java, the .jinit command of rJava needs to be called before loading rsubgroup, i.e.

library(rJava) .jinit(parameters="-Xmx2048M") # for two gigabytes heap space, for example library(rsubgroup)

Please note that this needs to happen before rJava is used in any way. After the JVM has been initialized (and started), setting the heap space has no effect any more. Therefore, it is recommended to execute the .jinit(...) command right after loading the rJava package.

Details

Author(s)

Martin Atzmueller

Maintainer: Martin Atzmueller <martin@atzmueller.net>

References

1. Martin Atzmueller and Frank Puppe. SD-Map - A Fast Algorithm for Exhaustive Subgroup Discovery. Knowledge Discovery in Databases: PKDD 2006, LNAI 4213, pp. 6-17, Springer Verlag, 2006.

2. Martin Atzmueller and Florian Lemmerich. Fast Subgroup Discovery for Continuous Target Concepts. In: Foundations of Intelligent Systems, LNCS 5722, pp. 35-44, Springer Verlag, 2009.

3. Florian Lemmerich and Mathias Rohlfs and Martin Atzmueller. Fast Discovery of Relevant Subgroup Patterns. In: Proc. 23rd FLAIRS Conference, AAAI Press, 2010.