Indian Fake Data Generator

A fast tool with no extra packages that creates realistic mock Indian profile data based on Census 2011 statistics.

Unlike other tools that make impossible combinations (like a Sikh named Mohammed Sharma from Mizoram), this tool links religion, state, caste, gender, and job together so the generated people make logical sense.

What It Can Do

• Linked Census Data: Correctly links Religion ➔ State ➔ Caste ➔ Gender ➔ Education ➔ Job ➔ Assets.
• Extra Features: Creates credit and health scores (with adjustable bias), builds a short text bio, and generates custom prompts for AI chat.
• Fast: Generates over 60,000 items per second.
• No Dependencies: Runs on standard Node.js without needing any other npm packages.

Installation

npm install @abhay557/indian-fakedata

Quick Start

1. Basic Generation

import { generate } from '@abhay557/indian-fakedata';

// Generate 10 standard profiles
const profiles = generate({ count: 10 });

// Generate with constraints
const filtered = generate({
  count: 1,
  constraints: { religion: 'Hindu', state: 'Tamil Nadu', gender: 'female' }
});

2. Exporting to a File

Generate extra credit/health stats and write them directly to a file:

import { generateEnriched, saveProfilesToFile } from '@abhay557/indian-fakedata';

const data = generateEnriched({
  count: 100,
  includeOutcomes: true,
  includeAgentPersona: true
});

// Save to disk (supports 'json', 'jsonl', or flat 'csv')
saveProfilesToFile(data, './output/data.csv', 'csv');

API Reference

generate(options?)

Creates a list of mock people.

interface GeneratorOptions {
  count?: number;                      // Default: 1
  seed?: number;                       // Number to get the exact same data every time
  constraints?: GenerationConstraints; // Filter by religion, state, caste, age, etc.
}

generateStream(options?)

For large amounts of data:

import { generateStream } from '@abhay557/indian-fakedata';

for (const profile of generateStream({ count: 1000000 })) {
  saveToDatabase(profile);
}

CLI Usage

You can also run the generator from your terminal using npx:

# Save 1000 people to a CSV file
npx Indian-FakeData -c 1000 -f csv -o output.csv

# Save 50000 people with extra features to a JSONL file
npx Indian-FakeData -c 50000 -f jsonl -o data.jsonl --enrich --bias 0.3

CLI Options

• -c, --count <n>: How many records to create (default: 100)
• -f, --format <fmt>: File format: json, jsonl, or csv
• -o, --output <path>: File path to save output
• -s, --seed <n>: Seed number for same results
• --enrich: Add credit, health, narrative, and AI prompt fields
• --bias <0-1>: Systemic bias level (default: 0.3)

License

MIT © Abhay Mourya