range.exponential / exponentialFrom #10

New issue

Closed

opened 2026-09-13 21:24:23 +09:00 by natsukium · 1 comment

natsukium commented

2026-09-13 21:24:23 +09:00

Owner

Missing

range.exponential min max and range.exponentialFrom origin min max.

Why

linear grows the bound proportionally to size, so with maxSize = 100 the span is at most 100x the smallest. Magnitudes like 2^53, 1.0e100 or "just below the Nix integer limit" are unreachable in a reasonable range. Hedgehog's Range.exponential* scale the bound geometrically with size; this is the only range constructor it has that nixcheck lacks.

Suggested shape

bounds size interpolates min + (max - min) * (2^(size*k/100) - 1) / (2^k - 1) or similar; ints and floats through the same formula (range.linear already works with float endpoints).
Best landed together with #3.

## Missing `range.exponential min max` and `range.exponentialFrom origin min max`. ## Why `linear` grows the bound proportionally to size, so with `maxSize = 100` the span is at most 100x the smallest. Magnitudes like `2^53`, `1.0e100` or "just below the Nix integer limit" are unreachable in a reasonable range. Hedgehog's `Range.exponential*` scale the bound geometrically with size; this is the only range constructor it has that nixcheck lacks. ## Suggested shape - `bounds size` interpolates `min + (max - min) * (2^(size*k/100) - 1) / (2^k - 1)` or similar; ints and floats through the same formula (`range.linear` already works with float endpoints). - Best landed together with #3.

natsukium referenced this issue

2026-09-13 21:24:52 +09:00

gen.float: a float generator with a range and edge values #3

natsukium commented

2026-09-14 00:44:14 +09:00

Author

Owner

Implemented in a043d6e, pushed to main.

Implemented in a043d6e, pushed to main.

natsukium closed this issue

2026-09-14 00:44:14 +09:00