RPHKEEP procedure

Saves information from a proportional hazards model fitted by RPHFIT (R.W. Payne).


Options

RESIDUALS = variate
Saves the standardized residuals

FITTEDVALUES = variate
Saves the fitted values

ESTIMATES = variate
Saves estimates of the parameters

SE = variate
Saves standard errors of the estimates

RESPONSE = variate
Saves the response variate defined for the generalized linear model

OFFSET = variate
Saves the offset variate defined for the generalized linear model

INDEX = variate
Index variate used to produce the expanded covariates and factors

RISKSET = factor
Saves the expanded time factor

_2LOGLIKELIHOOD = scalar
Saves -2 × log-likelihood for the fitted model


No parameters


Description

This procedure allows you to copy information into GenStat data structures from a proportional hazard model that has been fitted by procedure RPHFIT. You do not need to declare the structures in advance; GenStat will declare them automatically to be of the correct type and length.

   The RESIDUALS and FITTEDVALUES options save the standardized residuals and the fitted values. The ESTIMATES and SE options save the parameter estimates and their standard errors. The RESPONSE and OFFSET options save the response variate and the offset variate that have been defined for the generalized linear model. The INDEX variate saves the variate of indexes used to construct the expanded x-variates and factors from original variates and factors of the model. The RISKSET option saves a variate indicating the time interval corresponding to each of their units. Finally, the _2LOGLIKELIHOOD option saves -2 times the log-likelihood.

 

Options: RESIDUALS, FITTEDVALUES, ESTIMATES, SE, RESPONSE, OFFSET, INDEX, RISKSET, _2LOGLIKELIHOOD.

Parameters: none.


Method

The log-likelihood is calculated as described by Aitkin et al. (1989). The response variate and offset are recovered from a workspace structure that is defined by RPHFIT to hold details of the model. The other information is saved using RKEEP (which can also be used to save additional relevant output).


Reference

Aitkin, M., Anderson, A., Francis, B. & Hinde, J. (1989). Statistical Modelling in GLIM. Oxford University Press.