Random LSH Candidate Workload Generator

Description

Generate bounded string-set pair batches with exact supplied overlaps, then delegate banded MinHash candidate flags and signatures; retain exact Jaccard, estimated Jaccard, matching bands, and actual candidate counts separately, without interpreting a candidate as semantic similarity or a guaranteed retrieval outcome.

Synthetic statistical data separates a specified generating model from the values observed in a finite draw. A model parameter, a sample summary, and the result of a fitted analysis answer different questions; retaining them separately makes numerical and statistical checks easier to interpret.

When to use this tool

Use this generator to create reproducible lsh candidate workload cases from the supplied model or bank. Inspect the returned evidence before using cases in a test or exercise.

How to use the tool and how it works

Supply universe, pairCount, leftCount, rightCount, sharedCount, hashCount, bandCount in the field shapes below. Set seed when you need an explicit replay.
Run the tool and correct any input error. For generated controls, distinguish a successful fixture containing an invalid case from failure to generate the fixture.
Read settings, pairs, candidateCount. Use the conventions below and keep answer-bearing fields separate when presenting an exercise.

Construct set pairs with fixed overlap, compute exact and MinHash-estimated Jaccard similarity, and delegate banded locality-sensitive candidate selection. Return signatures, per-pair evidence, and candidate count.^[1]

Input field	Meaning and units	Accepted input
universe	Universe	Required. Array of Text; maximum UTF-16 length 64.; minimum items 1; maximum items 512.
pairCount	Pair Count	Required. Safe integer; minimum 1; maximum 32.
leftCount	Left Count	Required. Safe integer; minimum 0; maximum 128.
rightCount	Right Count	Required. Safe integer; minimum 0; maximum 128.
sharedCount	Shared Count	Required. Safe integer; minimum 0; maximum 128.
hashCount	Hash Count	Required. Safe integer; minimum 1; maximum 256.
bandCount	Band Count	Required. Safe integer; minimum 1; maximum 256.
seed	Optional numeric seed; takes precedence over context randomness. Without either, use the fixed tool.io default seed.	Optional. Safe integer.

Output field	Interpretation	Returned shape
settings	Settings	Required. Object; no unknown keys.
settings.hashCount	Hash Count	Required. Safe integer; minimum 1; maximum 256.
settings.bandCount	Band Count	Required. Safe integer; minimum 1; maximum 256.
settings.rowsPerBand	Rows Per Band	Required. Safe integer; minimum 1; maximum 256.
pairs	Pairs	Required. Array of Object; no unknown keys.; minimum items 1; maximum items 32.
pairs[].left	Left	Required. Array of Text; maximum UTF-16 length 64.; maximum items 128.
pairs[].right	Right	Required. Array of Text; maximum UTF-16 length 64.; maximum items 128.
pairs[].shared	Shared	Required. Array of Text; maximum UTF-16 length 64.; maximum items 128.
pairs[].exactJaccard	Exact Jaccard	Required. Finite number; minimum 0; maximum 1.
pairs[].estimatedJaccard	Estimated Jaccard	Required. Finite number; minimum 0; maximum 1.
pairs[].leftSignature	Left Signature	Required. Array of Safe integer; minimum 0; maximum 4294967295.; minimum items 1; maximum items 256.
pairs[].rightSignature	Right Signature	Required. Array of Safe integer; minimum 0; maximum 4294967295.; minimum items 1; maximum items 256.
pairs[].candidate	Candidate	Required. Boolean.
pairs[].matchingBands	Matching Bands	Required. Array of Safe integer; minimum 0; maximum 255.; maximum items 256.
candidateCount	Candidate Count	Required. Safe integer; minimum 0; maximum 32.
randomness	Replay metadata: source is seed or context; seed is the normalized 32-bit seed or null for a context source.	Required. Object; no unknown keys.
randomness.seed	Optional numeric replay seed; normalization and precedence are explained below.	Required. Safe integer. or Null.
randomness.source	Source	Required. Text; one of "context", "seed".

An explicit safe-integer seed takes precedence over injected context randomness and is normalized to an unsigned 32-bit integer. With no seed and no context source, replay uses the fixed tool.io default seed. A context source must return finite values from zero inclusive to one exclusive; its advancing state must be retained separately. randomness records the source and normalized seed, or null for context. These generators are not cryptographic random sources.

json

{
  "input": {
    "universe": [
      "a",
      "b",
      "c",
      "d"
    ],
    "pairCount": 2,
    "leftCount": 2,
    "rightCount": 2,
    "sharedCount": 1,
    "hashCount": 4,
    "bandCount": 4,
    "seed": 42
  },
  "output": {
    "settings": {
      "hashCount": 4,
      "bandCount": 4,
      "rowsPerBand": 1
    },
    "pairs": [
      {
        "left": [
          "a",
          "d"
        ],
        "right": [
          "b",
          "a"
        ],
        "shared": [
          "a"
        ],
        "exactJaccard": 0.3333333333333333,
        "estimatedJaccard": 0.5,
        "leftSignature": [
          2111704691,
          1185767313,
          1377171301,
          2111704691
        ],
        "rightSignature": [
          1185767313,
          1185767313,
          444641715,
          2111704691
        ],
        "candidate": true,
        "matchingBands": [
          1,
          3
        ]
      },
      {
        "left": [
          "d",
          "b"
        ],
        "right": [
          "a",
          "d"
        ],
        "shared": [
          "d"
        ],
        "exactJaccard": 0.3333333333333333,
        "estimatedJaccard": 0,
        "leftSignature": [
          1185767313,
          3736780160,
          444641715,
          3736780160
        ],
        "rightSignature": [
          2111704691,
          1185767313,
          1377171301,
          2111704691
        ],
        "candidate": false,
        "matchingBands": []
      }
    ],
    "candidateCount": 1,
    "randomness": {
      "seed": 42,
      "source": "seed"
    }
  }
}

Limitations and assumptions

hashCount must divide evenly into bandCount. Candidate status is an approximate retrieval filter, not exact similarity or guaranteed recall; hash work and universe size are bounded.
A replay requires the same complete input and implementation. Randomized construction does not guarantee uniform coverage of every feasible output; finite samples can repeat unless a distinct-selection rule is explicitly stated.
The field tables state structural bounds; cross-field constraints and bounded search or output budgets also apply. Unknown keys are rejected where the object contract disallows them. Input errors are distinct from null fields and deliberate invalid controls in a successful output.

Alternative or Complementary methods

For a supplied example, inspect the returned construction or delegated check directly. To evaluate a consumer, also test cases chosen for its own boundaries; a generated finite corpus does not establish exhaustive coverage.

References

Tool behavior and example: Random LSH Candidate Workload Generator — tool.io

Similar or alternative tools

Random MinHash Similarity Workload Generator
Generate seeded string sets with exact supplied left/right/shared cardinalities, reuse exact intersection/union and deterministic 32-bit MinHash signatures, and retain exact versus estimated Jaccard separately; empty-union Jaccard is one, and estimates are not guarantees or semantic-similarity judgments.
Random Heavy-Tail Comparison Experiment Generator
Compare matched-size Cauchy and Pareto Type I samples with visible derived sampler seeds, observed extremes, medians, and linearly interpolated empirical quantiles; preserve Pareto sampler 10-significant-digit rounding and epsilon tail clipping, and do not report a finite Cauchy population mean or variance.