Iteration History table for Fit Cox Model in a Counting Process Form

The iteration history displays the log-likelihood value at each step of the Newton-Raphson optimization algorithm that estimates the coefficients of the predictors. Step 0 of the iteration history is for a model with no terms where the coefficients, called starting or initial estimates, are all set to 0. At the last step, the coefficients are the values of the optimization algorithm parameters at convergence. If the algorithm fails to converge, then the coefficients are the values of the algorithm parameters at the last iteration. By default, the maximum number of iterations is 25. The log-likelihood values from the initial and final steps are the same log-likelihood values that are in the Model Summary table.