Is your feature request related to a problem? Please describe.
skpro currently has strong coverage of common continuous distributions, but it is missing a Gompertz distribution, which is widely used in survival analysis and actuarial modeling. This creates a gap for users modeling non-negative lifetime data with increasing hazard rates and forces them to use external wrappers instead of native skpro distributions.
Describe the solution you'd like
Add a new Gompertz distribution to skpro.distributions with scikit-compatible API and exact implementations for core methods:
-parameters: shape c and optional scale
-methods: _pdf, _log_pdf, _cdf, _ppf, _mean, _var, _energy_self, _energy_x
-proper broadcasting support, parameter validation, and tags (continuous, parametric, exact capabilities)
-export in distribution namespace and include in API reference docs
-add/extend tests so estimator checks and probabilistic method checks pass
Describe alternatives you've considered
Approximating with nearby distributions: possible in some cases but can be a poor fit when Gompertz hazard behavior is needed.
Creating only an adapter-level wrapper without full method implementations: faster to add but less explicit and potentially less maintainable.
Additional context
This is a natural addition alongside existing non-negative continuous distributions and is especially relevant to skpro’s survival focus. Assign this to me I can work on this.
Is your feature request related to a problem? Please describe.
skpro currently has strong coverage of common continuous distributions, but it is missing a Gompertz distribution, which is widely used in survival analysis and actuarial modeling. This creates a gap for users modeling non-negative lifetime data with increasing hazard rates and forces them to use external wrappers instead of native skpro distributions.
Describe the solution you'd like
Add a new Gompertz distribution to skpro.distributions with scikit-compatible API and exact implementations for core methods:
-parameters: shape c and optional scale
-methods: _pdf, _log_pdf, _cdf, _ppf, _mean, _var, _energy_self, _energy_x
-proper broadcasting support, parameter validation, and tags (continuous, parametric, exact capabilities)
-export in distribution namespace and include in API reference docs
-add/extend tests so estimator checks and probabilistic method checks pass
Describe alternatives you've considered
Approximating with nearby distributions: possible in some cases but can be a poor fit when Gompertz hazard behavior is needed.
Creating only an adapter-level wrapper without full method implementations: faster to add but less explicit and potentially less maintainable.
Additional context
This is a natural addition alongside existing non-negative continuous distributions and is especially relevant to skpro’s survival focus. Assign this to me I can work on this.