Comparison Based vs Non Comparison Based Sorting in DSA C++ - Trade-offs & Analysis

Mental Model

Sorting can be done by comparing elements or by using their values directly without comparing.

Analogy: Imagine sorting books by comparing their titles one by one (comparison based) versus sorting mail by zip codes using bins (non comparison based).

Unsorted array: [4, 2, 7, 1, 3]

Comparison based: compare pairs and swap -> gradually sorted
Non comparison based: use value properties (like digits) to place directly into buckets

Dry Run Walkthrough

Input: array: [4, 2, 7, 1, 3]

Goal: Sort the array in ascending order using both comparison based and non comparison based methods to see differences

Step 1: Start comparison based sort (Bubble Sort): compare 4 and 2, swap because 4 > 2

[2, 4, 7, 1, 3]

Why: We swap to move larger elements towards the end

Step 2: Compare 4 and 7, no swap needed

[2, 4, 7, 1, 3]

Why: 4 is less than 7, order is correct

Step 3: Compare 7 and 1, swap because 7 > 1

[2, 4, 1, 7, 3]

Why: Move larger element 7 towards the end

Step 4: Compare 7 and 3, swap because 7 > 3

[2, 4, 1, 3, 7]

Why: 7 moves to the last position

Step 5: Start non comparison based sort (Counting Sort): count occurrences of each number

Counts: 1->1, 2->1, 3->1, 4->1, 7->1

Why: Counting sort uses counts to place elements directly

Step 6: Rebuild sorted array by placing numbers in order based on counts

[1, 2, 3, 4, 7]

Why: Direct placement avoids comparisons

Result:

Sorted array: [1, 2, 3, 4, 7]

Annotated Code

DSA C++

#include <iostream>
#include <vector>
using namespace std;

// Comparison based sort: Bubble Sort
void bubbleSort(vector<int>& arr) {
    int n = arr.size();
    for (int i = 0; i < n - 1; i++) {
        for (int j = 0; j < n - i - 1; j++) {
            if (arr[j] > arr[j + 1]) {
                swap(arr[j], arr[j + 1]); // swap if out of order
            }
        }
    }
}

// Non comparison based sort: Counting Sort (works for positive integers)
void countingSort(vector<int>& arr) {
    if (arr.empty()) return;
    int max_val = arr[0];
    for (int val : arr) {
        if (val > max_val) max_val = val; // find max value
    }
    vector<int> count(max_val + 1, 0);
    for (int val : arr) {
        count[val]++; // count occurrences
    }
    int index = 0;
    for (int i = 0; i <= max_val; i++) {
        while (count[i] > 0) {
            arr[index++] = i; // place values in order
            count[i]--;
        }
    }
}

void printArray(const vector<int>& arr) {
    for (int val : arr) {
        cout << val << " ";
    }
    cout << "\n";
}

int main() {
    vector<int> arr1 = {4, 2, 7, 1, 3};
    vector<int> arr2 = arr1; // copy for second sort

    cout << "Original array: ";
    printArray(arr1);

    bubbleSort(arr1);
    cout << "After comparison based sort (Bubble Sort): ";
    printArray(arr1);

    countingSort(arr2);
    cout << "After non comparison based sort (Counting Sort): ";
    printArray(arr2);

    return 0;
}

if (arr[j] > arr[j + 1]) { swap(arr[j], arr[j + 1]); }

swap elements if current is greater to move larger elements right

for (int val : arr) { if (val > max_val) max_val = val; }

find maximum value to size count array

for (int val : arr) { count[val]++; }

count occurrences of each value

while (count[i] > 0) { arr[index++] = i; count[i]--; }

place values back into array in sorted order

Output

Original array: 4 2 7 1 3 

After comparison based sort (Bubble Sort): 1 2 3 4 7 

After non comparison based sort (Counting Sort): 1 2 3 4 7 

Complexity Analysis

Time: O(n^2) for comparison based (Bubble Sort) because it compares pairs repeatedly; O(n + k) for non comparison based (Counting Sort) where k is max value, because it counts and places directly

Space: O(1) extra space for comparison based (in-place swaps); O(k) extra space for non comparison based due to count array

vs Alternative: Comparison based sorts work on any data but are slower (at best O(n log n)); non comparison based sorts are faster for limited range data but need extra space and can't sort arbitrary data

Edge Cases

empty array

both sorts handle gracefully without errors

DSA C++

if (arr.empty()) return;

array with one element

array remains unchanged, no swaps or counts needed

DSA C++

for loops naturally skip or run once without issue

array with duplicates

both sorts correctly sort and keep duplicates

DSA C++

count[val]++ handles duplicates in counting sort

When to Use This Pattern

When sorting integers or data with limited range, consider non comparison based sorting for faster performance; otherwise, use comparison based sorting for general data.

Common Mistakes

Mistake: Trying to use counting sort on negative or non-integer data
Fix: Use comparison based sorting or adapt counting sort to handle negatives carefully

Mistake: Assuming non comparison based sorts always faster regardless of data
Fix: Remember non comparison based sorts depend on data range and type

Summary

Comparison based sorting uses element comparisons to order data; non comparison based sorting uses data properties like counts or digits.

Use comparison based sorting for general data and non comparison based sorting for integers with limited range for better speed.

The key insight is that non comparison based sorting avoids pairwise comparisons by leveraging data characteristics directly.