| Time comparison (100 000, 1000, lin) | Basic operations comparison (100 000, 1000, lin) |
|---|---|
 |
 |
sorting algorithms comparator that can be easily extended by new ones
tests all of the given sortings algorithms for consecutive sample sizes:
-
if mode is set to
lin(linear):1 *
intervalelements, 2 *intervalelements, 3 *intervalelements, ..., floor(max_sample_size/interval) elements -
if mode is set to
poly(polynomial):interval^1 elements,interval^2 elements,interval^3 elements, ..., floor(log(max_sample_size, base=interval)) elements -
if mode is set to
exp(exponential):2^(1 *
interval) elements, 2^(2 *interval) elements, 2^(3 *interval) elements, ..., floor(log(max_sample_size, base=2 ^ interval) elements
then generates line charts and .csv files containing data about algorithms' performance for every sample size
comparison is concerned in terms of time of execution and number of basic operations
repo contains also some sample data of default sorting algorithms' performance for each of modes
- python 3.5+
- matplotlib
just run the following command in your terminal
python run.py [max_sample_size] [interval] [mode (lin, poly, exp)]
as you can see, all of the arguments are optional and their default values are:
- max_sample_size = 10500
- interval = 750
- mode = lin
sorting algorithms used by default are bubble sort, selection sort, insertion sort, merge sort and quicksort
defaults can be changed in defaults.py module
create SortingTester instance (all parameters are optional and will be replaced by defaults if not provided)
then just call its run() and/or other methods
A folder containing:
- two .png files containing line charts for respectively time of execution and number of basic operations for each of sorting algorithms and every sample size
- two .csv files containing data about time of operation and number of basic operations for each algorithm and every sample size
custom sorting is a module containing a function implementing interface sort(list_to_sort: list) -> int
a custom sorting should sort input list and return number of basic operations performed
a custom sorting should be placed in sorting directory and the directory's __init__.py should be updated to export new sorting module