RPHCHANGE procedure

Modifies a proportional hazards model fitted by RPHFIT (R.W. Payne).


Options

PRINT = strings
Controls printed output (model, deviance, summary, estimates, correlations, fittedvalues, accumulated, monitoring, loglikelihood); default mode, summ, esti

METHOD = string
How to change the model (add, drop, switch); default add

POOL = string
Whether to pool terms in the accumulated summary generated by the fit


Parameter

TERMS = formula
Model specifying the change


Description

This procedure allows you to modify the contents of a proportional hazards model that has been fitted by procedure RPHFIT. The change to the model is specified by the TERMS parameter. The setting of the METHOD option specifies how the model is to be changed:

    add
adds the terms specified by the TERMS parameter to the fitted model;

    drop
drops those terms from the fitted model; and

    switch
drops any terms specified by the TERMS parameter that are already in the fitted model, and adds those that are not (i.e. this operates similarly to the SWITCH directive).

The default is METHOD=add. Note, though, that any term that is to be added must have been included in the full model specified by the MAXIMALMODEL option of RPHFIT. By default the changes are made individually, one term at a time, so that each one will have its own line in an accumulated analysis of deviance. However, you can set option POOL=yes to make them all at once.

   The PRINT option controls printed output with similar settings to those of the FIT directive, except that there is an extra setting loglikelihood to print -2 times the log-likelihood. The deviance produced for the terms in the regression model can be assessed using chi-square distributions as usual, but the residual deviance is not usable as the maximal model assumed by the generalized linear models method is inappropriate. So, the residual line is suppressed in the summary and accumulated analysis of deviance.

 

Options: PRINT, METHOD, POOL.

Parameter: TERMS.


Method

Further details of the method used here (and by RPHFIT) can be found in Aitkin et al. (1989).


Action with RESTRICT

None of the vectors must be restricted (and any restrictions will have been cancelled by RPHFIT).


Reference

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