- The separate Model and Results Summary Windows of @RISK 4.0 and 4.5 are now integrated into the Excel window.
- New @RISK Settings toolbar in Excel provides quick access to simulation settings.
- Graphs of simulation results and inputs can link directly to the cells they reference in Excel with “callout” windows.
- Nearly all modeling operations can be performed via drag and drop or simple clicks on the toolbar.
- Navigate through Results in Excel with <Tab> Key to move through @RISK inputs and outputs in open workbooks, with graphs pointing at the cell where the input or output is located.
- Support for versions of Microsoft Excel through Excel 2007, including the larger worksheet size of Excel 2007.
- All graphs and reports update live as a simulation runs
- Window selector allows tiling, minimizing, closing, activating etc. of all open @RISK windows
Model Window
- Distribution functions can be edited directly in the Model window
- Graph thumbnails in list graphically display input distributions
- Drag thumbnail graph off Model window to expand to full-size graph
- Statistics shown in table can be customized
- Report of Model Definition in Excel includes graph thumbnails
- Inputs are organized by categories, with default categories made based on row or column names in workbook
- Clicking or tabbing on an input in the table moves to relevant cell in Excel and displays graph
Adding Outputs
- Existing outputs and output ranges may be edited or deleted directly over Excel
- Properties dialog can be used to enter property functions
Correlations
- Correlations between distributions are quickly defined in matrices that pop-up over Excel
- Dynamic scatter plots of correlated inputs graphically display correlated samples between two inputs
- Correlation coefficients in matrix can be adjusted using slider
- Correlation Matrices defined in the @RISK interface can be located anywhere in Excel, not just on the @RISK Correlations sheet
- Correlated Time Series with Instances can be quickly set up to correlate sets of similar distributions which each use the same matrix
- Scatter plot matrix shows correlations generated between inputs in a matrix after a run
Distribution Fitting
- Data sets for fitting now reside in Excel
- Up to 10 overlays of different fitted distribution in a single graph
- Fit Manager tracks multiple fitted data sets in a model
Define Distributions
- Window stays up as you move to different cells to define or edit @RISK distribution functions
- <Tab> Key moves through cells in open workbooks and worksheets with distribution functions
- Edit functions directly in formula or in argument panel
- Up to 10 overlays of different distribution functions
- Graph settings and overlays maintained when returning to define distribution for a cell
- Separate argument panels for graph and overlays.
- Context sensitive (F1) help for each distribution function type.
- Properties dialog can be used to enter property functions
Simulation and Sampling
- Demo mode updates Excel, graphs and reports every iteration
- Eight random number generators available
- Auto-stop maximum iterations now 50,000
- New progress window reports additional run statistics and allows changing settings during run
- New Performance Monitor for @RISK Accelerator multi-CPU support
- Excel normal recalculation mode (standard or monte carlo) can be selected from the toolbar
- RiskStatic property function can store value to be returned from a distribution in a normal Excel recalculation
- Alternatively mean, mode, median or a percentile can be returned from a distribution in a normal Excel recalculation
Convergence Monitoring
- New algorithm based on user-defined tolerance and confidence interval settings, or on mean, Std. deviation, or percentile.
- User selectable statistics to test during monitoring
- Can be applied globally for all outputs or individually by output using the RiskConvergence property function
Results Window
- Graph thumbnails in list graphically display simulated distributions
- Drag thumbnail graph off Results window to expand to full-size graph
- Statistics shown in table can be customized
- Report of Results Window in Excel includes graph thumbnails
- Results window and thumbnail graphs update live as a simulation runs to show a “dashboard” of simulation results
- Clicking or tabbing on a result in the table moves to relevant cell in Excel and displays graph
Graphics - General
- New graphics engine, designed for processing simulation data, provides faster graphing and real-time animation of simulation results.
- Type Lower, Middle Range and Upper Delimiter Values directly on Graph
- Edit Titles directly on Graph
- User controllable number of delimiter bars for overlays
- Predefined and custom markers on graphs
- Extensive graph color and shading options
- Controllable number formatting on graphs
- Customizable statistics displayed in legend of graph
- Improved graphing of discrete distributions and data sets, including new RiskIsDiscrete property function
- Select overlays by clicking on cells to overlay in Excel
- Up to 10 overlays on a single graph
- Remove overlays by clicking on the entry in the legend
Scatter Plots
- Any pair of output results and/or sampled inputs can be placed in a scatter plot
- Overlay up to 10 scatter plots in a single chart
- X and Y values can be standardized
- User controllable Confidence Ellipse can be added
- Scatter plots can be generated in Excel as native Excel charts
Summary Graphs and Box Plots
- Any inputs or outputs (not just those in an output range) can be placed in a summary graph or box plot
- Box plot has user definable box and whisker settings
Sensitivity Analysis
- Pivoted Detailed Statistics report allows listings of larger numbers of inputs and outputs
- Regression – Mapped Values analysis type displays values, as opposed to coefficients, on bars
- Drag a bar from a Tornado Chart to show a scatter graph of values for a key input vs output
- Pre-screening inputs based on their precedence in formulas to outputs in your model helps keeps erroneous inputs from appearing in results.
- Sensitivity Scatter Plot Matrix can be displayed, showing a matrix of scatter plots ranked by most important inputs
- RiskMakeInput function allows sensitivity impacts of multiple inputs to be captured in a single bar in a tornado chart
Filtering
- Filtering can be specified using percentiles as well as actual values
- Filtering available directly from a graph window
- Percentile filters update real time as data changes during a simulation
@RISK Library
- The @RISK Library provides a repository for sharing @RISK inputs and simulation results.
- Distributions to be shared can be entered directly in the Library or added from an @RISK model in Excel
- Revision history is provided for library distributions
- Results for different simulations stored in the library can be graphically compared
- Archived simulations, including workbooks simulated, can be reloaded into Excel
Function Swap
- Function Swap allows @RISK functions to be removed and restored from workbooks; easing the sharing of workbooks with non-@RISK users.
- Value used to replace a swapped out distribution can be a user selected statistic from the distribution
- Model changes made when functions are swapped out are captured in the RiskStatic function
- Distributions swapped back in are placed in correct location in changed workbook
- Report on changes made to a swapped out model available
Data Reports
- Data from a simulation may be sorted to show key values you are interested in, including Error values
- Iterations from a previously run simulation may be stepped through, updating Excel with values sampled and results calculated.
Save Simulations
- Simulation Data and Graphs can be stored in the simulated workbook
- Simulation Data and Graphs, along with the simulated workbook, can be stored in the @RISK Library
Six Sigma Support
- 19 new statistics functions provide six sigma statistics in models
- LSL, USL and Target values set using new RiskSixSigma property function
- Results summary window displays six sigma statistics
- LSL, USL and Target values displayed with markers on graphs
- Six Sigma guide and help details how to develop six sigma models using @RISK
Clear Data
- Simulation results, settings, and/or all @RISK functions and information can be permanently removed from an Excel workbook
Application Settings
- Default settings for graphs, simulations, function swap, reporting and others are specified in a single location
- Standard settings can be shared between users with a custom configuration file
- Custom formatting of @RISK input and output cells, including color, font and borders
Excel Reports
- New standard headers and formatting
- Reports include graphics where applicable
- Ability to have more than one template sheet report.
Help and Tutorials
- Technical information on distribution functions integrated into help system
- On-line tutorials and movies illustrate key concepts
Installation, Licensing and Updates
- New installer and licensing system
- Licenses may be moved to a new computer via the internet
- Automatic updating of installations with new releases and patches
Programmability and Macros
- New object-oriented API has hundreds of properties and methods for setting up and running simulations and generating results from VBA
- Intellisense pops up available arguments and constants when wrting programs that control @RISK
New and Enhanced @RISK Functions in Excel
- RiskCompound, used for “frequency / severity” modeling, takes two distributions to form a new input distribution. RiskCompound requires two arguments, one for frequency and one for severity, and each an @RISK distribution function or a reference to a formula in Excel that generates the frequency or severity value. It also has two optional argument for insurance modeling – deductible and limit. Deductible, if entered, is subtracted from severity, and if deductible exceeds limit, severity is set to equal limit.
- RiskTheo functions such as RiskTheoMean or RiskTheoPtoX return a desired statistic on an input probability distribution.
- @RISK statistics functions for simulation results (such as RiskMean) and RiskTheo functions can take an optional RiskTruncate function to specify a percentile or actual range over which statistics should be calculated.
- RiskSensitivity function returns sensitivity analysis results directly to your spreadsheet.
- A new set of 19 @RISK statistics functions (such as RiskCPK) return a desired Six Sigma statistic on a simulation output.
- RiskConvergenceLevel returns the convergence level of a simulation output
- RiskStopRun stops a simulation when an Excel formula returns TRUE
- RiskMakeInput specifies that the calculated value for a formula will be treated as a simulation input, in the same manner as a distribution function. This function allows the results of Excel calculations (or a combination of distribution functions) to be treated as a single “input” in a sensitivity analysis.
- RiskResultsGraph supports overlays and graphs of input probability distributions
- RiskSimTable allows simulations to be named, and provide new icon in results viewing for managing output data.
New Property Functions for @RISK
Outputs and Distribution Functions
- RiskUnits specifies units to use to label an input or output in graphs and reports
- RiskSeed specifies the seed and random number generator to use for an individual distribution function
- RiskConvergence specifies specific convergence monitoring settings for an output
- RiskSixSigma specifies LSL,USL and Target values for an output
- RiskIsDiscrete specifies that the data set to be generated for an output should be treated as a discrete distribution when graphing and calculating statistics
- RiskStatic specifies the value to be returned by a distribution function in a normal Excel recalculation and optionally will automatically store the value in a formula that was replaced by a distribution function
- RiskTruncateP specifies truncation limits for a distribution function using percentile values
- RiskCategory property function to help organize inputs into functional groups.
- RiskLibrary is used to attach an input in the model to a probability distribution in the @RISK Library.