aggregateResults: Aggregation of Single BBoB Results
In andreas-he/cmaesbenchmarking: Benchmarking of cmaes Stopping Conditions

Description Usage Arguments Details Value

aggregateResults is used to aggregate the results generated by readOutput.

1	aggregateResults(allResults)

allResults

The user has to pass a number of result objects produced by applying readOutput.

The function aggregateResults takes a number of result objects, produced by applying readOutput on single data files, and aggregates those results. For example: The user can pass the results of two function optimization runs. First, he generates the results of each single output by calling readOutput for every single file (e.g. result1 = readOutput(output_function1), result2 = readOutput(output_function2). Then, the user passes these results to aggregateResults (e.g. aggregateResults(c(result1, result2))

aggregateResults returns a list that contains the following components aggregated over all passed single bbob results (see readOutput for information on the non aggregated single bbob results):

`aggregatedAllBest`	vector of all best fitness values.
`aggregatedAvgBest`	aggregated average best fitness values.
`aggregatedOverallBest`	overall best fitness values.
`aggregatedOverallWorst`	overall worst fitness values.
`aggregatedSDBests`	standard deviations of all best fitness values.
`aggregatedAllRuns`	vector of the number of iterations.
`aggregatedLongestRun`	overall highest required number of iterations.
`aggregatedShortestRun`	overall lowest required number of iterations.
`aggregatedAvgRun`	average required iterations.
`aggregatedSDRuns`	standard deviation of required iterations (`sd(aggregatedAllRuns`)).
`aggregatedAllRunsEval`	vector of the number of function evaluations.
`aggregatedLongestRunEval`	overall highest required number of function evaluations.
`aggregatedShortestRunEval`	overall lowest required number of function evaluations.
`aggregatedAvgRunEva`	average required function evaluations.
`aggregatedSDRunsEval`	standard deviation of required function evaluations (`sd(aggregatedAllRunsEval`)).
`aggregatedAllStagnation`	vector of the final iterations without improvement.
`aggregatedLongestStagnation`	overall highest number of iterations without improvement.
`aggregatedShortestStagnation`	overall number of iterations without improvement.
`aggregatedAvgStagnation`	average number of final iterations without improvement in the fitness value.
`aggregatedSDStagnation`	standard deviation of the number of final iterations without improvement (sd of `aggregatedAllStagnation`).
`aggregatedAllConvergence`	a `data.frame` that stores the average convergence of every single bbob result.
`aggregatedAvgConvergence`	the mean of the convergence of all single bbob results (`mean(aggregatedAllConvergence`)).
`aggregatedAllRestarts`	vector of the number of restarts.
`aggregated_t_test_termination`	vector of the number of t-test terminations while using OCD as a stopping condition.
`aggregated_chi_test_termination`	vector of the number of chi`^2`-test terminations while using OCD as a stopping condition.