This menu controls the actual solution of the problem. Solution is done in two stages. First, you build a model. In building the model, Parity Builder assigns fundamentally identical individuals to clusters. (Individuals involved in restrictions, and individuals with unique combinations of attribute values, will be assigned to their own personal clusters, called singletons.) Along the way, Parity Builder determines the range of values of each attribute in the population of individuals. (Until you generate a domain analysis or build the model, Parity Builder will not know what values exist for the attributes; until you build the model, Parity Builder will not know how those values are distributed among the individuals.) Once the model is built, you can tell Parity Builder to solve the model and produce assignments. As long as the model remains in memory, you can continue to issue solve commands (with or without changes to the solver settings, in the Settings tab), hoping to improve the quality of the solution.

Report domains
This instructs Parity Builder to scan all the individuals and determine what attribute values are present. Parity Builder writes an Attribute Domain Analysis to the Messages tab, listing the values it found for each categorical, numerical or affinity attribute (but not their distribution across the population of individuals).
Analyze
This option commands Parity Builder to analyze the distribution of each attribute across the population of individuals and to construct a model for the problem. Once the model is built, Parity Builder will write a Cluster Analysis in the Messages tab detailing how many clusters it found, as well as what values it saw for each non-label attribute (a list of values for each categorical or affinity attribute, a range of values for each numerical attribute) and their distribution across the individuals. Changes to anything other than solver settings will cause Parity Builder to delete the model.
Build rosters
This commands Parity Builder to start solving the model. It is inactive until a model is built. You can invoke the Build rosters command more than once with the same model (for instance, after aborting a run and changing settings, or after the run hits a time limit), in which case Parity Builder will resume solution while retaining the previous best known result. At the conclusion of a solution run (including one you abort), Parity Builder will write a candidate set of rosters to the Assignments tab, detailing which individuals are assigned to each roster, how large the roster is, the distribution of each categorical and affinity attribute value within the roster (with the distribution of that value across all individuals), and the mean value of each numerical attribute within that roster (again with the population mean for comparison purposes). Rosters can be exported to a comma-separated-value file using the Export rosters option on the File menu.
Abort
This option is available only while the solver is running. It tells Parity Builder to stop the solver as quickly as possible. Upon conclusion, the best known solution (if any) will be displayed in the Assignments tab. You can restart the solution process with the Build rosters option.
Find
Once a solution exists, use this option to locate the roster containing a particular individual. You will be presented with a list of all individuals. Select one and click OK. The Assignments tab will scroll until the roster containing that individual becomes visible. (Scrolling is a bit approximate. If you do not see the individual, scroll down a line or two; the individual may be just below the "horizon".)
Clear messages
This clears all messages written to the Messages tab, but leaves the model intact (so that you can resume solution if you wish).
Clear rosters
This clears the Assignments tab and deletes any solution that has been found.