We're pleased to introduce Version 2014 of Frontline Solvers for Excel, including a new release of our "super-product," Analytic Solver Platform. The biggest news in our V2014 release is the introduction of Dimensional Modeling, a powerful new facility for building what-if, optimization and simulation models that are better structured, far more compact, and easier to maintain and extend. But that's not all...
Better Interactive Performance
If you've used our software, you've been accustomed to the fastest Monte Carlo simulation in Excel, and the fastest available Solvers for mathematical optimization. But in V2014, we've focused on faster interactive performance, when starting up Excel, switching among worksheets and multiple open workbooks, and working with large models with thousands to hundreds of thousands of formulas, constraints and/or probability distributions. If you've ever found yourself waiting, we think you'll be pleased when you try V2014!
ProActive Technical Support
Active Support in our V12.5 release, with Live Chat to a support rep embedded in Excel, was a hit with our users. We've taken this a step further with ProActive Support. Our goal is to minimize time when you're "stuck", and bypass the initial steps of a tech support call or email, where users often struggle to describe what's wrong. In V2014, if you encounter a certain number of error messages or unusual conditions, we will automatically create a support ticket in our Help Desk, prompting our support reps to follow up with you by email.
As always, no information about your actual model, such as numbers or formulas in any worksheet cells, is ever reported to Frontline's servers. Only "events" that occur in our software, such as "an optimization was run", "the model was diagnosed as nonlinear", or "error message xyz appeared", are reported. And you can turn this on or off via the Help - Support Mode menu option. We are watching out for your privacy!
Dimensional Modeling in Analytic Solver Platform, Risk Solver Platform and Premium Solver Platform V2014 is a breakthrough capability for spreadsheet users. You can use it to build what-if models and perform sensitivity analysis in new ways. But you can also use it to build more structured, easily readable, compact optimization and simulation models, that represent business situations with many ‘dimensions’ such as products, customer types, geographic regions, or time periods.
If you’ve used Pivot Tables or PowerPivot in Excel, or if you’ve used a multidimensional database, data warehouse or business intelligence tool, the concepts behind Dimensional Modeling will probably be familiar. But these tools primarily deal with multidimensional data, generally reflecting the past. Dimensional Modeling makes it easy to construct multidimensional models, composed of formulas, that you can use in planning the future.
Use Everything You Know
You can use everything you know about Excel formulas in Dimensional Modeling. For example, if cell A1 contains a number and cells B2 and C3 contain =PsiCube() functions, a formula like =A1+B2*C3 in a cell will yield a “cube result,” multiplying and adding the appropriate values along several dimensions. You can also define B2 and C3 based on Pivot Tables containing input data, and easily obtain a Pivot Table that displays the results of =A1+B2*C3. Double-clicking a cube formula displays a built-in multi-dimensional cube result viewer with drag-and-drop pivoting of dimensions. Other Psi functions can write cube results to cell ranges.
Very often, a successful model is expanded to include more products, regions, time periods, etc. This is much easier to do with a Dimensional Model than with a traditional spreadsheet model. Formulas like =A1+B2*C3 need not change at all when you add elements (more cell or database values) to the existing dimensions. You can even add new dimensions to B2 and C3, or turn A1 from a single cell to a cube, and =A1+B2*C3 will automatically calculate results across the additional dimensions.
Better Optimization and Simulation Models
You can use Dimensional Modeling to create far more compact optimization and simulation models. For example, blocks of constraints that once required large ranges of repetitive formulas can now be entered in one or a few cells, drastically reducing the number of formulas. Similarly, in a simulation model, blocks of uncertain functions that once required large numbers of repetitive formula cells can be reduced to one or a few cells. You can define as many “structural dimensions” as you like, and use up to 8 of them at one time, in any calculated “cube intermediate result” in a formula.
Multi-Dimensional Parametric Analysis
With Dimensional Modeling, you can define as many parametric dimensions as you like, and create cubes of parameter values that are defined over one or more dimensions. You can then perform sensitivity analysis or run multiple optimizations or simulations, where any subset of your parametric dimensions are “active” (the others are “locked”), and get results for all combinations of the parameter values you’ve defined.
Just as you can expand a model by adding a structural dimension, automatically generalizing one or more cubes over the new dimension without changing your formulas, you can add a parametric dimension, use it in one or more cubes, and expand your sensitivity analysis, optimization or simulation runs to include more scenarios, without changing any formulas.
Some Assembly Required
If you’re used to conventional "columns and rows" spreadsheet modeling, Dimensional Modeling will require some learning and practice. But when you want to construct larger models for common business situations, the payoff from Dimensional Modeling can be huge. In our user survey one year ago, of all the possible enhancements we could offer, Dimensional Modeling was the idea that elicited the strongest positive response from our users. We look forward to seeing what you do with this powerful new facility!