Function for fitting proportioal hazard model with baseline hazard
Mstep(Y, R, X, B, Pen, lambda, cbx)An object with fields:
H = hazards (matrix),
cbx = coefficient estimates (vector),
lambda = proposal for new lambda,
ed = effective dimension,
G = G matrix,
ll = log-likelihood,
pen = penalized part of log-likelihood,
Mpen = penalized M matrix
Expected events (matrix)
Expected risk sets (matrix)
Covariates (matrix)
B-spline basis
Penalty matrix
Smoothing parameter (number)
Current coefficient estimates