Closed glemaitre closed 2 months ago
Could you please fix the shape information of the predict_cumulative_incidence docstring? It lacks the time dimension. I believe it should be (n_samples, n_events + 1, n_time_horizons)
I would do that in a subsequent PR.
There is a bug in
SurvivalBoost.predict_proba
. Thetime_horizon
parameter does not currently accept array-like while it should.