Given code:
[Test]
public void TestFoatsAndDoubles()
{
object x = 0.0500000007f;
Assert.That(x, Is.EqualTo(0.05).Within(0.0000001));
}
This boxes the floating point constant into an object to then do a numeric comparison.
The idea is to remove expensive boxing and subsequent unboxing when not needed.
Fixed code ideally should create:
[Test]
public void TestFoatsAndDoubles()
{
float x = 0.0500000007f;
Assert.That(x, Is.EqualTo(0.05).Within(0.0000001));
}