Map Predicted Hazard to Predicted Density for a Single Observation

Usage

map_hazard_to_density(hazard_pred_single_obs)

Arguments

hazard_pred_single_obs: A numeric vector of predicted hazard of failure in a given bin (under a given partitioning of the support) for a single observational unit based on a long format data structure (from format_long_hazards). This is the probability that a given value falls in a corresponding bin, given that it has not yet failed (fallen in a preceding bin), as per Díaz I, van der Laan MJ (2011). “Super learner based conditional density estimation with application to marginal structural models.” International Journal of Biostatistics, 7(1), 1–20. doi:10.2202/1557-4679.1356 . .

Value

A matrix composed of a single row and a number of columns specified by the grid of penalization parameters used in fitting of the highly adaptive lasso. This is the predicted conditional density for a given observation, re-mapped from the hazard scale.

Details

For a single observation, map a predicted hazard of failure (as an occurrence in a particular bin, under a given partitioning of the support) to a density.