pub const EPSILON: f64 = 1e-10;
Epsilon value for floating-point comparisons
Used to handle floating-point precision errors. Two floats are considered equal if their absolute difference is less than EPSILON.