Synthetic Data Values Generator (SDVG)

Language

• English
• Русский

Description

SDVG (Synthetic Data Values Generator) is a tool for generating synthetic data. It supports various run modes, data types for generation, and output formats.

Run modes:

• CLI - generate data, create configs, and validate them via the console;
• HTTP server - accepts generation requests through an HTTP API.

Data types:

• strings (english, russian);
• integers and floating-point numbers;
• dates with timestamps;
• UUID.

String subtypes:

• random strings;
• texts;
• first names;
• last names;
• phone numbers;
• patterns.

Each data type can be generated with the following options:

• specify percentage/number of unique values per column;
• ordered generation (sequence);
• foreign key reference;
• idempotent generation using a seed number;
• value generation from ranges with percentage-based distribution.

Output formats:

• devnull;
• CSV files;
• Parquet files;
• HTTP API;
• Tarantool Column Store HTTP API.

Quick Start

Here's an example of a data model that generates 10,000 user rows and writes them to a CSV file:

output:
  type: csv
models:
  user:
    rows_count: 10000
    columns:
      - name: id
        type: uuid
      - name: name
        type: string
        type_params:
          logical_type: first_name

Save this as simple_model.yml, then run:

./sdvg generate simple_model.yml

This will create a CSV file with fake user data like id and name:

id,name
c8a53cfd-1089-4154-9627-560fbbea2fef,Sutherlan
b5c024f8-3f6f-43d3-b021-0bb2305cc680,Hilton
5adf8218-7b53-41bb-873d-c5768ca6afa2,Craggy
...

To launch the generator in interactive mode:

./sdvg

To view available commands and arguments:

./sdvg -h
./sdvg --help
./sdvg generate -h

More information can be found in the user guide.

Documentation

• User Guide
• Developer Guide
• Goals and Standards Compliance
• Changelog
• License

Maintainers

• @hackallcode
• @ReverseTM
• @choseenonee
• @Hoodie-Huuuuu