Fig. 8
From: A strategy for optimal fitting of multiplicative and additive hazards regression models
Log-transformed smoothed pseudo-observations against the continuous covariate age for 9 deciles of the event time distribution
Fig. 8 | BMC Medical Research Methodology