BenchDesign: Create a new BenchDesign object

Description Usage Arguments Value Author(s) See Also Examples

Description

Initializes a new BenchDesign object of benchmarking methods and data.

The BenchDesign class serves as the core container for methods and data used for benchmarking in the SummarizedBenchmark package. The object can be initialized with a list of methods ot be benchmarked, a default benchmarking data set, both or neither. Methods must be passed to the constructor as BDMethod or BDMethodList objects.

The constructor can also be used to access the BenchDesign stored in a SummarizedBenchmark object.

Usage

1
2
3
4
5
6
7
BenchDesign(..., methods = NULL, data = NULL)

## S4 method for signature 'ANY'
BenchDesign(..., methods = NULL, data = NULL)

## S4 method for signature 'SummarizedBenchmark'
BenchDesign(methods, data)

Arguments

...

named set of BDMethod objects and/or unnamed BenchDesign objects. Only the methods of any BenchDesign object will be used, and the data slot of the objects will be ignored.

methods

named set of BDMethod objects and/or unnamed BenchDesign objects as a list. (default = NULL)

data

optional data.frame or other list object to be used in the benchmark. (default = NULL)

Value

BenchDesign object.

Author(s)

Patrick Kimes

See Also

BenchDesign-class, BDMethod, BDMethodList

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
## with no input
bd <- BenchDesign()

## with data - data must be a named argument
datadf <- data.frame(pval = runif(20), x1 = rnorm(20))
bd <- BenchDesign(data = datadf)

## with two methods and data
method_bh <- BDMethod(stats::p.adjust, params = rlang::quos(p = pval, method = "BH"))
method_bf <- BDMethod(stats::p.adjust, params = rlang::quos(p = pval, method = "bonferroni"))
bd <- BenchDesign(bh = method_bh, bonf = method_bf,
                  data = datadf)

## with BDMethodList and data
bdml <- BDMethodList(bh = method_bh, bonf = method_bf)
bd <- BenchDesign(methods = bdml, data = datadf)

SummarizedBenchmark documentation built on Nov. 8, 2020, 8:30 p.m.