Fixed example and reproducibility for Log-likelihood

CamDavidsonPilon / lifelines

Survival analysis in Python

lifelines.readthedocs.org

MIT License

2.32k stars 551 forks source link

Fixed example and reproducibility for Log-likelihood #1547

Closed Genarito closed 11 months ago

Genarito commented 11 months ago

Fixes #1545: a model with the highest Log-Likelihood is the best one. Also, random_state prevents Pandas random sampling from generating a counterexample to what the documentation states (sometimes the "best model" could be the lower score)