Content of this site
Installation How to install ReLife on you machine
Basics ReLife basic things to know before starting
API The content of the exposed ReLife API
Why you should use ReLife ?
Statiscal expertise

ReLife implements approved statistical models of reliability and renewal theory

Open science

ReLife is open source and tries to gather the best of scientific knowledge to conduct appropriate risk analysis for asset management

Easy to use

ReLife is coded in Python and was designed to be accessible for user from any programming background or experience level

They use ReLife
RTE
Artelys
Company 3
Company 4
Release notes
  • 2.0.0 version is a significant refactored version of the unique ReLife release. It might be better for contributions, testing, and long-term maintainability.
  • Parametric models, including lifetime models and stochastic processes, are now unified under one ParametricModel interface which includes a composite parameter structure, making model model composition and reuse easy
  • The new version makes use of extensive type hinting, including generic abstract base classes. It enhances modularity and clarifies object interaction logic. Currently, the codebase is type hint in the source files. It adds a lot of boilerplate code. So we plan to pass it to stub files (see below) in next releases
  • ReLife is built on NumPy. Thus, in v2.0.0, we try to incluse rigorous testing for array shape validation and broadcasting compatibility, ensuring consistent function behavior.
  • The current release proposes a new documentation. The old one was buggy and not maintained. The new document is still incomplete. Writting is long and it will be done continuously. In the long term, we hope that it will make the API more accessible, including tutorials, mathematical explanations, and usage examples to encourage collaboration
🚀 Ongoing work
  • Enhance type hinting with mypy validations. We also plan to experiment with stub files to consolidate type hints in a single location, improving code readability.
  • Adding more documentation, notebook tutorials, high-level math explanations of ReLife functionnalities, a contribution guide explaning conception choices
  • Implementing Non Homogeneous Poisson Process -> v2.1.0
  • Implementing Gamma Process -> v2.2.0
  • Implementing Cox Regression -> v2.3.0
  • Creating an Orange Data Mining pluggin to have a GUI -> v(?)
  • Adding low-level bindings to speed up some code (renewal equations) : possibilities are cython or pybind11 -> v(?)