summarizeExperiments: Summarize selected experiments.

In BatchExperiments: Statistical Experiments on Batch Computing Clusters

Summarize selected experiments.

Description

A data.frame is returned that contains summary information about the selected experiments. The data.frame is constructed by building the columns “prob, <prob.pars>, algo, <algo.pars>, repl”. Now only the columns in show will be selected, how many of such experiments exist will be counted in a new column “.count”.

Usage

summarizeExperiments(reg, ids, show = c("prob", "algo"))

Arguments

reg	[ExperimentRegistry] Registry.
ids	[integer] Selected experiments. Default is all experiments.
show	[character] Should detailed information for each single experiment be printed? Default is c("prob", "algo").

Value

[data.frame].

Examples

reg = makeExperimentRegistry("summarizeExperiments", seed = 123, file.dir = tempfile())
p1 = addProblem(reg, "p1", static = 1)
a1 = addAlgorithm(reg, id = "a1", fun = function(static, dynamic, alpha, beta) 1)
a2 = addAlgorithm(reg, id = "a2", fun = function(static, dynamic, alpha, gamma) 2)
ad1 = makeDesign(a1, exhaustive = list(alpha = 1:2, beta = 1:2))
ad2 = makeDesign(a2, exhaustive = list(alpha = 1:2, gamma = 7:8))
addExperiments(reg, algo.designs = list(ad1, ad2), repls = 2)
print(summarizeExperiments(reg))
print(summarizeExperiments(reg, show = c("prob", "algo", "alpha", "gamma")))

BatchExperiments documentation built on March 21, 2022, 5:06 p.m.