Skip to content

Latest commit

 

History

History
36 lines (23 loc) · 991 Bytes

README.md

File metadata and controls

36 lines (23 loc) · 991 Bytes

hypervector

hypie

Status: 🔨 Pre-first release

Hypervector is a test data fixture engine intended for data-intensive production systems such as those reliant on machine learning and data science components.

Use

Hypervector sits naturally alongside your integration and smoke test infrastructure during continuous integration stages, and makes large high-dimensional feature vectors available via endpoint.

The service allows for benchmarks to be saved alongside their parent feature vectors. These can be asserted by test runners via the Hypervector API, providing a sophisticated regression testing framework for statistical properties of your fixtures.

Build

You can run the test suite and build a local binary (useful for development) with:

make all

Building and running a Docker image is done with:

make docker