Sorting: your code, counted
Dynamic programming
Backtracking
Greedy
Divide and conquer
Graphs
Exam
Write your sort (or start from a template) and see how much work it really does: comparisons, reads and writes, on arrays of 8 to 1024 elements.
Text & documents expand_more
Numbers & maths expand_more
Data & formats expand_more
Security expand_more
Development & DevOps expand_more
Artificial Intelligence expand_more
Finance expand_more
Health & Wellness expand_more
Productivity expand_more
Games & Entertainment expand_more
Multimedia & design expand_more
Business expand_more
Your code, counted
Write a function sort(a) that sorts the array a in place, or start from a template. It runs in your browser, apart from the page and with an 8-second limit, on arrays of 8 to 1024 elements that are already sorted, reversed and random. Your code does not need any changes to be counted: every time you compare two elements with <, > or by subtracting them, that is one comparison; every a[i] is a read and every a[i] = … a write.
How it grows
With the large sizes, the tool tests which curve best fits your counts: n, n log n or n², and with what constant (“≈ 0.5·n²”). It is a measurement, not a proof: an algorithm can look n log n at these sizes and not be.
Stability
A sort is stable if two elements with the same key stay in the order they came in. The tool checks it with an array full of repeated keys that remember their original position.
What is not counted
Comparisons with === (they do not call valueOf) and whatever you do in auxiliary arrays, except when you write back into a. Comparing an element with a plain number counts as half a comparison.
