make_test_ped: Creation of a generic test PED dataframe

In pkopper/deeppam: Machine Learning for Time-to-Event Data with Semi-Structured Deep Piecewise Additive Models

Description

This function creates a list that contains (possibly) necessary PED frames for prediction. The list features three elements: [[1]]: the PED frame using the original end times (original=. (Could be used for validation loss) [[2]]: the PED frame using the maximally observed end time for all observations (transformed). (Could be used for testing) [[3]]: the raw (no PED) data (raw_data). The fucntions uses pammtools::as_ped in the backend.

Usage

make_test_ped(data, ped_train)

Arguments

data	a data.frame (or variant) for which the test set(s) should be computed.
ped_train	a PED data set (ee pammtools package) that has been used for training

Value

a list of length 3. (See description)

Author(s)

Philipp Kopper

pkopper/deeppam documentation built on Jan. 19, 2021, 12:39 a.m.