US Brand Signal Co-occurrence Analyzer

US Brand Signal Co-occurrence Analyzer measures how often buyer-supplied opaque signal IDs appear together in bounded observation groups. Give it groups of IDs; it returns stable signal counts, unordered pair counts, support, Jaccard overlap, and SHA-256 digests.

The Actor only analyzes the JSON you provide. It does not fetch a website, verify a signal, identify a person or company, infer causality, or make an external truth claim.

What you submit and receive

You submit 1 to 100 groups. Each group has a printable ASCII groupId and 1 to 30 unique printable ASCII signalIds. The Actor sorts group and signal IDs before analysis, so input order does not change the result.

You receive one Dataset report containing:

• signalCounts: how many submitted groups contain each signal.
• pairRows: stable unordered pairs with cooccurrenceCount, support, and jaccard.
• groupCounts and reconciled totals.
• SHA-256 digests for the input, groups, signal counts, pair rows, and complete result.
• A truthful OUTPUT receipt after one confirmed Dataset write.

For a pair, cooccurrenceCount is the number of groups containing both IDs. support is that count divided by the number of groups. jaccard is the count divided by the union of groups containing either ID. Values are rounded to six decimal places for stable JSON.

Run it

1. Open Try for free in the Actor Console.
2. Paste the exact JSON shape shown in the Input section.
3. Start the run and read the one report row from the Dataset.
4. Read the terminal delivery receipt from OUTPUT.

The complete runnable prefill is also available in examples/input.json and public-task.json.

Pricing

Pay per event: $0.005 per run start plus $0.010 per delivered result on the FREE tier. The result event is emitted only after one confirmed Dataset write. The Actor checks that the buyer cap covers both charges before processing.

Tier	Discount	Start	Delivered result
FREE	0%	$0.00500	$0.01000
BRONZE	5%	$0.00475	$0.00950
SILVER	10%	$0.00450	$0.00900
GOLD	15%	$0.00425	$0.00850
PLATINUM	18%	$0.00410	$0.00820
DIAMOND	20%	$0.00400	$0.00800

Invalid input, insufficient cap, uncertain delivery, and an uncharged or multiply charged result are not reported as successful results. The Actor never retries a paid Dataset push.

Freshness and determinism

There is no external freshness because this Actor does not fetch data and does not read the clock. Refresh the groups in your upstream system and rerun when observations change. The same groups always produce the same rows and digests, regardless of input order.

Input

{
  "schemaVersion": "1.0",
  "groups": [
    {"groupId": "observation-001", "signalIds": ["signal-a", "signal-b", "signal-c"]},
    {"groupId": "observation-002", "signalIds": ["signal-a", "signal-b"]},
    {"groupId": "observation-003", "signalIds": ["signal-a", "signal-c"]},
    {"groupId": "observation-004", "signalIds": ["signal-b", "signal-c"]}
  ]
}

Output

The Dataset has one deterministic report item. pairRows is sorted by signalA, then signalB. A pair is stored only once with its IDs in ASCII order. OUTPUT contains the delivery status, receipt, and links represented by the Store output schema.

Related public Actors

These Actors can feed or use the same buyer-owned signal workflow. This Actor does not call them.

Actor	Use it for
US Brand Signal Metrics Aggregator	Summarize individual signal rows before comparing groups.
US Brand Signal Brief Composer	Turn a submitted signal report into a bounded brief.
US Brand Action Queue	Convert selected signals into next-action rows.
US Brand Evidence Snapshot Diff	Compare two buyer-supplied evidence snapshots.
US Brand Offer Evidence Normalizer	Normalize offer evidence before downstream analysis.

Limitations

• IDs are opaque and are not treated as names, people, companies, or verified entities.
• The maximum input is 100 groups and 30 unique IDs per group.
• Duplicate group IDs or duplicate IDs within a group fail closed.
• The report is bounded at 48 KiB. A mathematically valid input with too many distinct pairs can return output_too_large instead of a partial result.
• There is no network, proxy, BYOK key, LLM, clock, state, child Actor, or Dataset input.
• JSON Schema validates shape; runtime validators additionally enforce ordering, digest, and cross-field reconciliation.

FAQ

Does co-occurrence prove a relationship? No. It is only a count in the groups you submitted.

What is support? The fraction of submitted groups containing both members of a pair.

What is Jaccard overlap? The pair's shared-group count divided by the groups containing either member.

Can I submit names or URLs? The contract accepts opaque ASCII IDs only. The Actor does not resolve identity or fetch URLs.

Why did a valid-looking run return output_too_large? The bounded report protects buyers from an incomplete or unexpectedly expensive result. Reduce the number of groups or IDs per run, or split the input upstream.

What happens after uncertain delivery? The Actor records a non-success terminal receipt and does not issue a second Dataset write.

Built by zinin.