Integer slopes inside a prescribed rational interval #
Choose a fixed field block width and dimension, then reserve a strict gap between the allowed copy exponent and projective-direction capacity. The prefix fraction can still lie in any prescribed interval above that copy exponent. This is an exact integer construction, without limiting notation.
Numerical slopes supporting a geometric scheduler and a prescribed upper bound on the fraction of input bits used as source prefixes.
- dimension : ℕ
Affine-space dimension, fixed independently of the input length.
- blockWidth : ℕ
Bits per field block on the selected field-size subsequence.
- copySlope : ℕ
Scheduled batch exponent per complete field block.
- inputSlope : ℕ
Input bits per complete field block.
Every field block has positive width.
The affine space has positive dimension.
The common-zero-block code admits the selected dimension.
Strict exponential room absorbs the scheduler's fixed packing constant.
The prefix leaves a positive suffix slope.
The scheduled exponent strictly exceeds the requested exponent.
- prefixRate : upperDenominator * (self.dimension * self.blockWidth + 1) < upperNumerator * self.inputSlope
The prefix fraction is below the chosen coefficient threshold.
Instances For
Every nonempty rational interval above the copy exponent contains the prefix fraction of a suitable geometric parameter choice.