readOutput: Interpretation of BBoB Data
In andreas-he/cmaesbenchmarking: Benchmarking of cmaes Stopping Conditions

Description Usage Arguments Details Value

readOutput is used to read and interpret an output file generated by the benchmarking functions bbob_custom or bbob_custom_parallel.

1	readOutput(file)

file

The user has to specify the exact datapath to the bbob data to be read and interpreted.

There are several functions for reading and interpreting the output generated by bbob_custom or bbob_custom_parallel. readOutput is used to read one single output file and returns a list of results of that data (see value section). The user might execute something as follows: readOutput(output_function1).

readOutput returns an object of the class single_bbob_result. A list of those can be passed to aggregateResults. Each single_bbob_result is a list that contains the following components:

`allBest`	best fitness value of each processed function instance.
`avgBest`	average best fitness value of all processed function instances.
`overallBest`	overall best fitness value of all processed function instances.
`overallWorst`	overall worst fitness of all processed function instances.
`sdbest`	standard deviation of all best fitness values (`sd(allBest)`).
`allRuns`	number of iterations of each processed function instance that has been required for optimization.
`longestRun`	overall highest required number of iterations of all function instances.
`shortestRun`	overall lowest required number of iterations of all function instances.
`avgRuns`	average required iterations of all functions instances.
`sdRuns`	standard deviation of required iterations of all function instances (`sd(allRuns`)).
`allRunsEval`	number of function evaluations of each processed function instance that has been required for optimization.
`longestRunEval`	overall highest required number of function evaluations of all function instances.
`shortestRunEval`	overall lowest required number of function evaluations of all function instances.
`avgRunEval`	average required function evaluations of all functions instances.
`sdRunsEval`	standard deviation of required function evaluations of all function instances (`sd(allRunsEval`)).
`allStagnations`	number of final iterations without improvement in the fitness value.
`longestStagnation`	highest number of iterations without improvement in the fitness value of all function instances.
`shortestStagnation`	lowest number of iterations without improvement in the fitness value of all function instances..
`avgStagnation`	average number of final iterations without improvement in the fitness value.
`sdStagnation`	standard deviation of the number of final iterations without improvement in the fitness value of all function instances (`sd(allStagnation`)).
`allConvergence`	a `data.frame` that stores the gap between the fitness value and the global optimal value for each iteration.
`avgConvergence`	a vector that stores `mean(allConvergence)` of all processed instances applied over all iterations.
`allRestarts`	number of restarts occured in optimizing one single function instance.
`t_test_termination`	number of t-test terminations while using OCD as a stopping condition.
`chi_test_termination`	number of chi`^2`-test terminations while using OCD as a stopping condition.
`functionID`	the identifier of the bbob function that has been optimized.
`dimension`	the problem dimensions for the optimization.