Module vector_set 

Element and set operations over the integer vector carriers.

The float vector nodes (super::vector_math) compute over embeddings; these read integer vectors as lists: how many elements, what is at an index, does a value appear, where, and how two lists compare as sets. A host holding a list of ordinals — the records matching a label, the queries matching a predicate — asks these the questions it would otherwise ask a counter.

That is the point of them. A rank (“how many matching records come before this one”) and a cardinality (“how many match at all”) look like things a counter produces as records go by, but both are pure functions of the list and the record: the rank is the position, the cardinality is the length. Written that way the answer is owned by its inputs, replays from its coordinates, and agrees on every engine and every fiber, where a counter’s answer depends on how many times it was called (Runtime Model §9.1, “No state across pulls”).

Each operation exists for vec_i32 and vec_i64, since the carrier is part of the wire’s type and polydat does not silently widen one to the other. An index or a value that is out of range answers rather than panicking, because a miss is an ordinary result here: vec_position returns the length when the value is absent, which is “past the end” and the same convention vec_len bounds.

Structs

StrToVecF64

str_to_vec_f64(s) — a JSON array of numbers as a vec_f64 wire, such as a window of samples for is_stable. Fails by name on anything that is not a JSON array of numbers.

StrToVecI32

str_to_vec_i32(s) — a JSON array of integers as a vec_i32 wire.

StrToVecI64

str_to_vec_i64(s) — a JSON array of integers as a vec_i64 wire.

VecAtI32

vec_at_i32(v, i) — the element at i, or 0 past the end.

VecAtI64

vec_at_i64(v, i) — the element at i, or 0 past the end.

VecContainsI32

vec_contains_i32(v, x) — 1 when x appears in the list, else 0.

VecContainsI64

vec_contains_i64(v, x) — 1 when x appears in the list, else 0.

VecCountBelowI32

vec_count_below_i32(v, x) — how many elements are less than x.

VecCountBelowI64

vec_count_below_i64(v, x) — how many elements are less than x.

VecIntersectCountI32

vec_intersect_count_i32(a, b) — how many distinct values appear in both lists.

VecIntersectCountI64

vec_intersect_count_i64(a, b) — how many distinct values appear in both lists.

VecLenI32

vec_len_i32(v) — how many elements the list holds.

VecLenI64

vec_len_i64(v) — how many elements the list holds.

VecMaxI32

vec_max_i32(v) — the largest element, or 0 for an empty list.

VecMaxI64

vec_max_i64(v) — the largest element, or 0 for an empty list.

VecMinI32

vec_min_i32(v) — the smallest element, or 0 for an empty list.

VecMinI64

vec_min_i64(v) — the smallest element, or 0 for an empty list.

VecPositionI32

vec_position_i32(v, x) — the index of the first x, or the list’s length when it does not appear.

VecPositionI64

vec_position_i64(v, x) — the index of the first x, or the list’s length when it does not appear.

VecSetEqI32

vec_set_eq_i32(a, b) — 1 when both lists hold the same set of values, whatever their order or repetition.

VecSetEqI64

vec_set_eq_i64(a, b) — 1 when both lists hold the same set of values, whatever their order or repetition.