Recipe catalog / comparable-fit

Check whether a comparable property fits the subject

Is comparable similar enough to subject in type, size, age, condition, and location wording to support a valuation comparison?

You need to screen candidate comparable sales or rentals against a subject property from their descriptions, so that only plausible comparables reach a valuation model or an appraiser's review.

Explore this recipe interactively · Source and implementation guide

Use comparable-fit in TypeScript

Install with npm install jev-recipes. Requires Node.js 22.9 or newer and ES modules. Set TYPESAFE_API_KEY in your server environment for live calls, which send input to TypeSafe and use API quota. See the installation guide.

import { comparableFit } from 'jev-recipes/comparable-fit';

const result = await comparableFit({
  "subject": "Single-family ranch, 3 bedrooms, 2 baths, 1,650 sq ft, built 1978, updated kitchen and roof (2019), attached 2-car garage, 0.3-acre lot on a quiet suburban cul-de-sac in Maple Grove.",
  "comparable": "Condominium, 2 bedrooms, 2 baths, 1,100 sq ft, built 2015, 14th floor of a downtown high-rise with concierge, one deeded parking space, HOA dues $610/month.",
  "minConfidence": 0.8
});
console.log(result);

Input contract

{
  "input": {
    "$schema": "https://json-schema.org/draft/2020-12/schema",
    "type": "object",
    "properties": {
      "subject": {
        "type": "string"
      },
      "comparable": {
        "type": "string"
      },
      "minConfidence": {
        "type": "number",
        "minimum": 0,
        "maximum": 1
      }
    },
    "required": [
      "subject",
      "comparable"
    ]
  },
  "result": {
    "$schema": "https://json-schema.org/draft/2020-12/schema",
    "type": "object",
    "properties": {
      "model": {
        "type": "string"
      },
      "usage": {
        "type": "object",
        "properties": {
          "input_tokens": {
            "type": "integer",
            "minimum": 0,
            "maximum": 9007199254740991
          },
          "output_tokens": {
            "type": "integer",
            "minimum": 0,
            "maximum": 9007199254740991
          }
        },
        "required": [
          "input_tokens",
          "output_tokens"
        ],
        "additionalProperties": false
      },
      "status": {
        "type": "string",
        "enum": [
          "ready",
          "review"
        ]
      },
      "probability": {
        "type": "number",
        "minimum": 0,
        "maximum": 1
      },
      "confidence": {
        "type": "number",
        "minimum": 0,
        "maximum": 1
      },
      "verdict": {
        "type": "string",
        "enum": [
          "comparable",
          "dissimilar"
        ]
      }
    },
    "required": [
      "model",
      "usage",
      "status",
      "probability",
      "confidence",
      "verdict"
    ],
    "additionalProperties": false
  }
}

Saved example result

This hand-authored response demonstrates the contract. It is not a model accuracy measurement. Run it without an API key: npx jev-recipes demo comparable-fit.

{
  "model": "demo-fixture",
  "usage": {
    "input_tokens": 0,
    "output_tokens": 0
  },
  "status": "ready",
  "probability": 0.04,
  "confidence": 0.96,
  "verdict": "dissimilar"
}

Evaluation evidence

Use the evaluation guide to measure this decision on your own labeled cases.

Limitations

• Judges described attributes only, not market data. Sale prices, dates, distances, and adjustments belong in valuation code.
• Does not know your tolerance bands for size, age, or distance; it applies a general standard of similarity, so encode specific thresholds in code.
• A comparable can fit on described attributes and still be a poor choice because of a stale sale date or unusual sale conditions the description omits.

Related recipes

• entity-match: Use entity-match to decide whether two records describe the same property, rather than whether two different properties are similar enough to compare.
• passage-compare: Use passage-compare to pick which of two passages better answers a question, rather than to judge similarity between two property descriptions.