You must have the Enterprise version of Anaplan to use Optimizer, as Optimizer is an entitled service. When Optimizer is in your Customer, or Tenant, it'll be available in all workspaces within the Customer or Tenant.
Your Customer Success Business Partner enables Optimizer in your model. Once enabled, Customer Success can help you set up Optimizer in your enabled model. Although your solution will be specific to your business needs and constraints, the process is the same for any Optimizer implementation. You configure an Optimizer action, and then create pages to gather inputs and present results.
Configure the Optimizer action
Optimizer is a model action. In the action, you define the three elements common to every optimization problem:
- Objective. This is what you want to maximize or minimize. For example, profit or cost. For this, you define the objective function. This specifies how the solution's quality is measured. It must be a numeric line item with a linear formula depending on the variables.
- Variables. The decisions you want to optimize. For example, how much to produce or where to allocate resources. These are the unknowns that Optimizer will find the best set of values for to meet the objective. Variables must be numeric line items. Summary must be None.
- Constraints. Business rules and limitations that Optimizer must account for. For example, capacity limits or minimum requirements. Constraints must be Boolean line items with linear formulas depending on the variables. The only comparative operators supported in expressions are greater than or equal to (>=), equal to (=), and less than or equal to (<=). Summary must be All.
When you add an Optimizer action, the Optimizer dialog collects this detail.
The Optimizer dialog
| Field | Description |
| Button text | The text that'll appear on the button that you associate with this action. |
| Time Out | The number of seconds that the optimization can run for before stopping. 4 hours (14400 seconds) is the maximum. Settings higher than this are ignored. |
| Problem | Choose between Linear Programming and Feasibility Programming. Linear programming looks for the optimal solution. Feasibility programming looks for a feasible solution. This can be useful when testing, as proving feasibility generally takes less time that finding the optimum. |
| Objective | Choose between Minimize and Maximize. |
| Line Item | Select the line item that has the objective function. This is what the optimization will try to minimize or maximize. |
| Tolerance % (MIPGap) | The tolerance within which optimality is declared and results returned. The default is 0.01. |
| Import on timeout | If selected, import the results on time-out even if they don't meet the specified tolerance. |
| Variable | Select the numeric line items that represent the variables in the optimization. |
| Constraint | Select the Boolean line items that represent the constraints that the optimization must account for. |
Configure a page
To use your Optimizer action, typically you will develop a board or worksheet with:
- Input boxes where end users can enter values needed for your scenario.
- A button that runs your Optimizer action.
- Variable fields to display your optimized solution.
This enables any end user with access to your Optimizer dashboard to update the inputs and run the Optimizer action as often as needed.
See Optimizer limitations for more details on how an Optimizer action should be configured and implemented.
Automation
Automation of an action that runs the Optimizer is supported by Anaplan Connect and Anaplan API.
