SA with features apart from time

class ProportionalHazard[source]

ProportionalHazard(breakpoints:array, t_scaler:MaxAbsScaler, x_scaler:StandardScaler, dim:int, h:tuple=(), **kwargs) :: Module

Hazard proportional to time and feature component as shown above. parameters:

  • breakpoints: time points where hazard would change
  • max_t: maximum point of time to plot to.
  • dim: number of input dimensions of x
  • h: (optional) number of hidden units (for x only).

Fitting Cox Proportional Hazard Model

Plotting hazard functions

model.baseλ.plot_hazard()
x = df.drop(['t', 'e'], axis=1).iloc[4]
t = np.arange(df['t'].max())
model.plot_survival_function(t, x)