Estimation

From the Algorithm pulldown, select the run method.

For individual models:

Naive pooled

For population models, choose from:

FOCE ELS (FOCE Extended Least Squares)

QRPEM (Quasi-Random Parametric Expectation Maximization)

Laplacian

FOCE L-B (First-Order Conditional Estimation, Lindstrom-Bates)

FO (First Order)

IT2S-EM (Iterated two-stage expectation-maximization)

Naive pooled (only choice for Visual Predictive Check and Simulation run modes)

For more on run methods, see “Model Engines”.

From the ODE Solver pulldown, select one of the ODE (ordinary differential equations) solver methods.

For more on ODE methodology, see “Differential equations in NLME” in the Phoenix NLME documentation.

Enter the Maximum Number of Iterations to use with each modeling run in the field.

Check the Sort Input checkbox to sort the input by subject and time values.

When checked, the data is automatically sorted by ID (up to 5 levels of ID) and then by Time (if the model is time-based). Unchecking this option will process the data in the order given in the input dataset, and thus if the same subject ID is present in the dataset but the records are not consecutive they will be treated as two different subjects.

If a single ID entry within a column is non-numeric then the entire column is considered non-numeric, otherwise it is considered numeric. Note that non-numeric and numeric ID variables are sorted differently. For example, in numeric ID sorting, 2 comes before 10 but in non-numeric order 10 comes before 2.

The Sort Column menu allows you to select input columns to use for sorting the output.

Once a column is selected, another Sort Column menu is displayed for selecting an additional column.