This tool performs benchmarks for some classical sorting and search algorithms taught in almost every algorithm analysis course. These benchmarks aim to empirically prove/demostrate the computational complexity of those algorithms.
For the empirical analysis and computational time complexity estimation of the algorithms, the benchmarks keep track of the number of times the basic operation (BO) of the corresponding algorithm is performed over the given input.
The benchmarks generate a data file containing the performance measuraments of the chosen algorithm for the provided input vector, as well as another file with the actual output of the benchmark. More precisely, the benchmark computes the minimum, maximum and average number of performed basic operations, and the total clock execution time of the algorithm over the given input.
The benchmarking tool can be used as
$ ./benchmark <options>The following is a list of the available options in the tool.
| Option | Description |
|---|---|
--method <name> |
Specifies the method to be benchmarked. |
--min-permutation-size <integer> |
Minimum permutation size to be tested. |
--max-permutation-size <integer> |
Maximum permutation size to be tested. |
--n-permutations <integer> |
Number of permutations used to compute the average benchmark parameters for a fixed permutation size. |
--delta-permutation-size <integer> |
Increment in the size of the permutations in each iteration. |
--dump-file <filepath> |
File to dump the benchmark data to be analised. |
The following is a list of the supported algorithms with their corresponding computational time complexity. For the sake of completeness, the space complexity has also been included to distinguise in-place algorithms from those that need auxiliary memory.
| Algorithm | Benchmark name | Worst time complexity | Average time complexity | Best time complexity | Space complexity |
|---|---|---|---|---|---|
| Bubblesort (naive) | naive-bubblesort |
||||
| Bubblesort | bubblesort |
The MIT License. See LICENSE for more information.