genesynth 0.2.1

Python package: genesynth

[](https://github.com/sterling312/genesynth/actions/workflows/github-actions-unittests.yaml)

# genesynth Library to synthetically generate data for declarative data structures.

# install ` pip install -r requirements.txt `

# example ` >>> from genesynth.orchestration import * >>> pipe = Orchestration.read_yaml('tests/graph.yaml') >>> pipe.run() >>> asyncio.run(pipe.root.save('graph.csv')) `

` $ pip install genesynth $ python -m genesynth.cli -f tests/test.yaml --stdout `

# project status

## supported feature * load yaml as configuration file * arbitrary row size support * data type mapping with configurable parameters * JSON (semi-structured data) support * improved data type support * foreign relationship support * DOT file graph * table graph * built-in orchestrator using graph * thread and process support * intermediary data temporary cache * graph visualization

## key features to add * add yaml validator * fix header support * additional output file formats (JSON, PSQL dump, CSV with quotes, etc) * add support for JSON arrays * improve constraint support * add support for quoted string * add support statistical distribution via kernel convolution * optimize orchestration and disk cache efficiency * optimize thread/process based generation * convert serial to autoincrement constraint for integer type * convert password to constraint of string type

## nice to have features to add * support external scheduler * support NLP based text generation * support sklearn * support integration with pytorch embedding * support for object reference via $ref * fix compatibility with [JSON schema array notation](https://json-schema.org/understanding-json-schema/reference/array.html#items) * fix when json array child object appears as separate items