When constructing a HashSet, does the capacity argument take the load factor into account?

The .NET API gives the following remarks for the HashSet<T>(Int32) constructor:

Since resizes are relatively expensive (require rehashing), this attempts to minimize the need to resize by setting the initial capacity based on the value of the capacity [argument].

The phrase based on the value gives me faith that the constructor does something smarter than simply setting the underlying array capacity to capacity. But will I be guaranteed to be able to add capacity items to the set with no resize?

The underlying load factor is of course an implementation detail. I checked if it is exposed by any API method, and it apparently isn’t. So hopefully .NET is using its knowledge of the load factor the way I want it to in this constructor.

>Solution :

The implementation can be seen on GitHub by clicking the source link in the docs.

After skipping through a few method calls, and a guard where it throws if capacity < 1, we get to the interesting part.

/// <summary>
/// Initializes buckets and slots arrays. Uses suggested capacity by finding next prime
/// greater than or equal to capacity.
/// </summary>
private int Initialize(int capacity) {
  int size = HashHelpers.GetPrime(capacity);

So, it does not directly use the load factor in the calculation, but using the next prime instead is still a pretty desirable implementation, as discussed here.