pub fn assert_approx( actual: impl Into<f64>, expected: impl Into<f64>, epsilon: impl Into<f64>, )
Assert that two floating point values are approximately equal. Accepts both f32 and f64 via Into so callers don’t need manual casts.