Hash Table Load Factor

The load factor of a hash table is the ratio of the number of elements stored in the hash table to the size of the underlying array (number of buckets). It is a measure of how full the hash table is and can have a significant impact on the performance of the hash table operations.

Load Factor = Number of Elements / Number of Buckets

For example, if a hash table has 10 elements and an array size of 20, the load factor would be 0.5 (10 / 20).

The load factor affects the performance of a hash table in the following ways:

Collision Resolution:

• As the load factor increases, the probability of collisions also increases.
• With more collisions, the collision resolution mechanism (e.g., separate chaining or open addressing) has to work harder to resolve them.
• This leads to longer chains or more probing steps, which can degrade the performance of lookup, insertion, and deletion operations.

Memory Usage:

• A lower load factor means that the hash table has more empty buckets, resulting in wasted memory space.
• On the other hand, a higher load factor may lead to more collisions but utilizes memory more efficiently.

Rehashing:

• When the load factor exceeds a certain threshold (e.g., 0.75), the hash table may need to be resized to maintain its performance.
• Resizing involves creating a new larger array and rehashing all the elements to redistribute them.
• Rehashing is an expensive operation that can impact performance, especially if it occurs frequently.

Here's an example to illustrate the effect of load factor on performance:

1#include <chrono>
2#include <iostream>
3#include <unordered_map>
4
5int main() {
6  using namespace std::chrono;
7
8  // Create hash tables with different load factors
9  // Low load factor
10  std::unordered_map<int, int> map1(1000);
11
12  // High load factor
13  std::unordered_map<int, int> map2(100);
14  
15  // Insert elements into the hash tables
16  for (int i = 0; i < 1000; i++) {
17    map1[i] = i;
18    map2[i] = i;
19  }
20
21  // Measure lookup time for map1
22  auto start1 = high_resolution_clock::now();
23  for (int i = 0; i < 10000; i++) {
24    map1.find(i);
25  }
26  auto end1 = high_resolution_clock::now();
27  auto duration1 =
28    duration_cast<microseconds>(end1 - start1);
29
30  // Measure lookup time for map2
31  auto start2 = high_resolution_clock::now();
32  for (int i = 0; i < 10000; i++) {
33    map2.find(i);
34  }
35  auto end2 = high_resolution_clock::now();
36  auto duration2 =
37      duration_cast<microseconds>(end2 - start2);
38
39  std::cout << "Lookup time with low load factor: "
40    << duration1.count() << " microseconds\n";
41  std::cout << "Lookup time with high load factor: "
42    << duration2.count() << " microseconds\n";
43}

In this example, we create two hash tables with different load factors - map1 with a low load factor and map2 with a high load factor. We insert the same number of elements into both hash tables and measure the lookup time for each.

The output might resemble:

1Lookup time with low load factor: 15 microseconds
2Lookup time with high load factor: 28 microseconds

The actual values may vary depending on your system, but the general trend is that the lookup time is higher for the hash table with a high load factor due to more collisions and longer collision resolution.

In practice, it's important to choose an appropriate load factor that balances performance and memory usage. A common rule of thumb is to keep the load factor below 0.75 to maintain good performance. When the load factor exceeds this threshold, the hash table is typically resized to a larger capacity to reduce collisions and improve performance.