The ICPHREG Procedure

The "Model Information" table displays the two-level name of the input data set, the type of model fitted, the parameterization type that is used for the baseline function, and the name and label of the dependent variables. If you specify the FREQ statement, this table also displays the name and label of the frequency variable.

If you specify a CLASS statement, PROC ICPHREG outputs the "Class Level Information" table. This table displays the design information for the classification variables when the parameterization is full rank.

If you specify a CLASS statement, PROC ICPHREG outputs the "Class Variable Levels" table. This table displays the levels of classification variables when the parameterization is not full rank.

The "Fit Statistics" table displays the negative of twice the log likelihood, Akaike’s information criterion (AIC), the corrected Akaike’s information criterion (AICC), and the Bayesian information criterion (BIC). This table is displayed by default.

The "Analysis of Maximum Likelihood Parameter Estimates" table displays the parameter name, the degrees of freedom for each parameter, the maximum likelihood estimate of each parameter, the estimated standard error of the parameter estimator, confidence limits for each parameter, a chi-square statistic for testing whether the parameter is 0, and the associated p-value for the statistic. This table is displayed by default.

If you specify the ITHISTORY option in the PROC ICPHREG statement, the procedure outputs a table that contains the following for each iteration in the iterative procedure for model fitting: the iteration number, the negative of twice of the log likelihood, the gradient, and values of all parameters in the model.

If you specify the ITHISTORY option in the PROC ICPHREG statement, the procedure displays the last evaluation of the gradient vector.

If you specify the ITHISTORY option in the PROC ICPHREG statement, the procedure displays the last evaluation of the Hessian matrix.

If you specify the COVB option in the MODEL statement, the procedure displays the estimated covariance matrix. This matrix is defined as the inverse of the information matrix at the final iteration and is based on the Hessian matrix that is used at the final iteration.

If you specify the CORRB option in the MODEL statement, the procedure displays the estimated correlation matrix, which is based on the Hessian matrix that is used at the final iteration.

If you specify the HAZARDRATIO statement, PROC ICPHREG outputs the "Hazard Ratios for label" table. The table displays the estimate and confidence limits for each hazard ratio. The ODS name of the "Hazard Ratios for label" table is HazardRatios.

If you specify the ITHISTORY, COVB, or CORRB option in the MODEL statement, PROC ICPHREG outputs the "Parameter Information" table. This table displays the names of the parameters and the corresponding level information of effects that contain the CLASS variables. The ODS name of the "Parameter Information" table is ParmInfo.

If a piecewise constant model is used (the default model), PROC ICPHREG outputs the "Constant Hazard Time Intervals" table. This table displays information about the parameterization of the baseline function under the piecewise constant model. The ODS name of the "Constant Hazard Time Intervals" table is HazardParms.

If you fit a cubic spline model, PROC ICPHREG outputs the "Cubic Spline Parameters" table. This table displays information about the parameterization of the baseline function under the cubic spline model. The ODS name of the "Cubic Spline Parameters" table is HazardParms.