pub fn integer_limit<F: Float>(t: impl Tolerant) -> f64Expand description
The magnitude below which a whole number compares with integers as it does exactly, where each integer converts to a float of F’s width and compares within tolerance t: whole_limit(t), and at most 2^p, below which floats of the width hold every whole number. A residue of one integer by another, where their magnitudes add up to less than it, is exact at the width, with or without a fused multiply-add.