picorules-compiler-js-eadv-mocker

Mock EADV data generator for Picorules testing. Generates synthetic clinical data based on compiled ruleblocks.

Installation

npm install picorules-compiler-js-eadv-mocker

Usage

Basic Usage

import { generateMockData } from 'picorules-compiler-js-eadv-mocker';

const result = generateMockData({
  ruleblocks: [
    {
      name: 'ckd',
      text: `
        egfr_last => eadv.lab_bld_egfr.val.last();
        has_ckd : {egfr_last < 60 => 1}, {=> 0};
      `,
      isActive: true,
    }
  ]
});

console.log(result.eadv);
// [
//   { eid: 1001, att: 'lab_bld_egfr', dt: '2024-12-01', val: 45.2 },
//   { eid: 1001, att: 'lab_bld_egfr', dt: '2024-11-15', val: 52.8 },
//   ...
// ]

With Custom Options

const result = generateMockData({
  ruleblocks: [...],
  options: {
    entityCount: 5,              // Generate 5 patients
    entityIdStart: 2000,         // Start IDs at 2000
    observationsPerEntity: 4,    // 4 observations per attribute
    dateRange: {
      start: '2023-01-01',
      end: '2024-12-31',
    },
    seed: 12345,                 // Reproducible random data
  }
});

With Custom Value Generators

import {
  generateMockData,
  clinicalValueGenerators,
  createRangeGenerator
} from 'picorules-compiler-js-eadv-mocker';

const result = generateMockData({
  ruleblocks: [...],
  options: {
    valueGenerators: {
      // Use built-in clinical generators
      ...clinicalValueGenerators,
      // Or create custom ones
      'my_custom_attr': createRangeGenerator(0, 100, 1),
    },
  }
});

With Bind Dependencies

For ruleblocks that reference other ruleblocks via bind statements:

const result = generateMockData({
  ruleblocks: [
    {
      name: 'monitoring',
      text: `
        ckd => rout_ckd.ckd.val.bind();
        egfr => eadv.lab_bld_egfr.val.last();
        needs_monitoring : {ckd = 1 and egfr < 45 => 1}, {=> 0};
      `,
      isActive: true,
    }
  ],
  options: {
    bindTableValues: {
      'rout_ckd': {
        'ckd': () => Math.random() > 0.5 ? 1 : 0,
      }
    }
  }
});

console.log(result.routTables);
// {
//   'rout_ckd': [
//     { eid: 1001, ckd: 1 },
//     { eid: 1002, ckd: 0 },
//     { eid: 1003, ckd: 1 }
//   ]
// }

API

generateMockData(input)

Main entry point. Parses ruleblocks and generates mock data.

Parameters:

• input.ruleblocks - Array of ruleblock inputs
• input.options - Optional MockerOptions

Returns: MockDataResult

generateMockDataFromParsed(parsed, options)

Generate mock data from already-parsed ruleblocks.

Options

| Option | Type | Default | Description | |--------|------|---------|-------------| | entityCount | number | 3 | Number of patients to generate | | entityIdStart | number | 1001 | Starting entity ID | | observationsPerEntity | number | 3 | Observations per attribute per entity | | dateRange | object | Last year | Date range for observations | | dateFormat | string | 'iso' | Date format: 'iso', 'oracle', 'mssql' | | dateDistribution | string | 'uniform' | Date distribution mode (see below) | | valueGenerators | object | {} | Custom generators per attribute | | defaultValueGenerator | function | random 0-100 | Default value generator | | includeMockBindTables | boolean | true | Generate rout_* tables for binds | | bindTableValues | object | {} | Custom generators for bind tables | | seed | number | Date.now() | Random seed for reproducibility |

Date Distributions

The mocker supports three date distribution modes to simulate realistic clinical patterns:

'uniform' (default)

Dates are evenly distributed across the date range with some jitter.

|-------|-------|-------|
   ×        ×       ×

'recent-weighted'

More observations cluster toward recent dates using exponential decay. Simulates typical clinical scenarios where recent visits are more frequent.

const result = generateMockData({
  ruleblocks: [...],
  options: {
    dateDistribution: 'recent-weighted',
  }
});

|-------|-------|-------|
                × ×× ×××

'clustered'

Observations group into 1-3 episodes, simulating hospital stays or illness episodes where multiple tests are ordered together.

const result = generateMockData({
  ruleblocks: [...],
  options: {
    dateDistribution: 'clustered',
  }
});

|-------|-------|-------|
   ×××            ××

Built-in Value Generators

The package includes clinical value generators for common attributes:

import { clinicalValueGenerators } from 'picorules-compiler-js-eadv-mocker';

// Available generators:
// - lab_bld_egfr (15-120)
// - lab_bld_haemoglobin (80-180)
// - lab_bld_hba1c (4-12)
// - lab_bld_glucose (3-20)
// - lab_bld_potassium (3-6)
// - ... and many more

License

MIT