Reference

Creating Objects

Several classes are defined in the package, most importantly, the BenchDesign and SummarizedBenchmark classes for encapsulating the benchmark design and benchmark results, repectively. The following set of functions can be used to create new instances of each class. For more details, see class definitions below.

BenchDesign()

Create a new BenchDesign object

BDMethod()

Create a new BDMethod object

BDMethodList()

Create a new BDMethodList object

BDData()

Create a new BDData object

SummarizedBenchmark()

Create a new SummarizedBenchmark object

Executing Benchmarks

Benchmark experiments can be executed from either BenchDesign or SummarizedBenchmark objects using the following set of functions.

buildBench()

Execute BenchDesign

updateBench()

Check/Update SummarizedBenchmark

BenchDesign Methods

Methods and data for benchmark experiments are stored in BenchDesign objects. The following functions are available for working with these objects to add, remove, or modify methods in the design, as well as to compare the methods and data contained in these objects.

addMethod()

Add method to BenchDesign object

dropMethod()

Remove method from BenchDesign object

expandMethod()

Expand method in BenchDesign object

modifyMethod()

Modify method in BenchDesign object

printMethod() printMethods()

Pretty print methods in a BenchDesign object

compareBenchDesigns()

Compare BenchDesign objects

compareBDMethod()

Compare BDMethod objects

compareBDData()

Compare BDData objects

hashBDData()

Hash data in BDData object

SummarizedBenchmark Methods

Results from benchmark experiments are stored in SummarizedBenchmark objects. The following functions are available for working with these objects to define benchmark metrics and evaluate or plot the results of a benchmark experiment.

groundTruths()

Get ground truths in SummarizedBenchmark object

availableMetrics()

List pre-defined metrics for SummarizedBenchmark objects

performanceMetrics()

Get performance metrics in SummarizedBenchmark object

addPerformanceMetric()

Add performance metric to SummarizedBenchmark object

estimateMetricsForAssay() estimatePerformanceMetrics()

Estimate performance metrics in SummarizedBenchmark object

tidyUpMetrics()

Tidy up performance metrics in SummarizedBenchmark object

plotROC()

Plot ROC curve for SummarizedBenchmark object

plotMethodsOverlap()

Plot UpSetR for SummarizedBenchmark object

Class Definitions

Two core classes, BenchDesign and SummarizedBenchmark are defined in the package, along with several smaller classes which help encapsulate data used in benchmarking (BDData) and definitions of methods benchmarked (BDMethod, BDMethodList).

BenchDesign-class

BenchDesign class

BDData-class

BDData class

BDMethod-class

BDMethod class

BDMethodList-class

BDMethodList class

SummarizedBenchmark-class

SummarizedBenchmark class

Class Setters

In addition to using the methods defined above for working with BenchDesign objects and working with SummarizedBenchmark, some slots and values of the classes can be directly modified using the following functions.

`BDData<-`()

Set data in BenchDesign object

`BDMethod<-`()

Set method in list or BenchDesign object

`BDMethodList<-`()

Set method list in BenchDesign object

`performanceMetrics<-`()

Set performance metrics in SummarizedBenchmark object

`groundTruths<-`()

Set ground truths in SummarizedBenchmark object

`mcols<-`(<SummarizedBenchmark>)

Set meta data columns in SummarizedBenchmark object

`assayNames<-`(<SummarizedBenchmark>,<character>)

Set assay names in SummarizedBenchmark object

Data Sets

Example data sets are included to illustrate the functionality of the package in the various vignettes and case studies.

allSB

SummarizedBenchmark object of isoform quantification results

sb

SummarizedBenchmark example

tdat

Example data.frame containing results for 50 two-sample t-tests.