Users of any of the DecisionTools Suite products will be notified automatically of any updated releases with a pop up message that once clicked will open a dialog window with a variety of options to update the software. You can also check for available updates by clicking on Help>Check for Software Updates in the @RISK ribbon.
Build #104 - July 12, 2016
New and Improved Tornado Graphs
Input Shading, Tornado Overlays and Contribution to Variance tornado graphs have all been added to the already powerful collection of tornado graphs that previous versions of @RISK had to offer.
Input Shading (See #1 in image below)
We've added a shading option to our Change in Output Mean Tornado graphs to have the ability to quickly see whether the input associated with each bar is high or low when the output statistic increases or decreases. In the example below you can see that when inputs such as Product Lifetime and Initial Unit Price are high, there is a positive impact on the net present value (NPV) of the project; when an input such as Initial Cost is high it will have a negative impact on the NPV.
Contribution to Variance (See #2 in image below)
@RISK's new Contribution to Variance tornado graphs help you understand how much of the variance in the output variable is attributable to each individual input. There are options for displaying both the magnitude and direction of the bars, or just the magnitude. The former displays bars to the left and right of the centerline, depending on the correlation between the input and the output, while the latter displays all input bars to the right (as shown below), so the contribution to variance can be more easily compared.
Tornado Overlays (See #3 in image below)
You can now overlay multiple tornado graphs to make specific comparisons easy to understand and communicate to others. This is especially helpful when you want to compare pre-mitigation vs post-mitigation on the same model; or if you would simply like to compare the results of multiple simulations to analyze other strategies.
RISKOptimizer is 4x Faster
RISKOptimizer now utilizes multiple CPUs (or cores) to dramatically speed up optimizations. Speed tests show optimizations run four times faster than before – or more – saving you tremendous time!
New Probability Distribution Functions
Functions are at the heart of risk analysis involving Monte Carlo simulation. In version 7.5 we have added a total of 16 new distribution functions. The new distribution functions will appeal to a variety of industries and applications, ranging from insurance risk to reliability engineering to modeling of household income. These new functions are important for accurate, insightful estimation of uncertainty.
- RiskBurr12(γ, β, α1, α2) specifies a Burr12 distribution with the parameters γ, β, α1 and α2. Also known as the Singh-Maddala distribution, the RiskBurr12 distribution is commonly used to model household income, insurance risk, and reliability data.
- RiskCauchy(γ, β) specifies a Cauchy distribution with the parameters γ and β. Often used as an example of a “pathological” distribution since it has no mean, variance, or higher moments, the RiskCauchy distribution is commonly used in scientific and engineering applications to model resonance behavior in matter, measurement repeatability, and light-dispersion.
- RiskCauchyAlt(arg1type, arg1value, arg2type,arg2value) specifies a Cauchy distribution with two arguments of the type arg1type and arg2type. These arguments can be either “gamma”, “beta”, or a value between 0 and 1 to specify a percentile.
- RiskDagum(γ, β, α1, α2) specifies a Dagum distribution with the parameters γ, β, α1 and α2. The RiskDagum distribution is most commonly used in the modeling of income distribution and actuarial statistics.
- RiskFatigueLife(γ, β, α) specifies a FatigueLife distribution with the parameters γ, β, and α. Also called the Birnbaum-Saunders distribution, the RiskFatigueLife distribution is commonly used to model crack propagation, the failure of materials over time, and related concepts.
- RiskFatigueLifeAlt(arg1type, arg1value, arg2type,arg2value , arg3type,arg3value) specifies a RiskFatigueLife distribution with three arguments of the type arg1type, arg2type and arg3type. These arguments can be either “gamma”, “beta”, "alpha" or a value between 0 and 1 to specify a percentile.
- RiskFrechet(γ, β, α) specifies a Frechet distribution with the parameters γ, β, and α. The Fréchet distribution is commonly used to model extreme events. Used in hydrology to model peak annual rainfall, damn overflow, and related concepts.
- RiskFrechetAlt(arg1type, arg1value, arg2type,arg2value , arg3type,arg3value) specifies a Frechet distribution with three arguments of the type arg1type, arg2type and arg3type. These arguments can be either “gamma”, “beta”, "alpha" or a value between 0 and 1 to specify a percentile.
- RiskHypSecant(γ, β) specifies a HypSecant distribution with the parameters γ, and β. The hypersecant distribution is quite similar to the normal distribution, but has a larger kurtosis, giving it a sharper peak.
- RiskHypSecantAlt(arg1type, arg1value, arg2type,arg2value) specifies a HypSecant distribution with two arguments of the type arg1type and arg2type. These arguments can be either “gamma”, “beta” or a value between 0 and 1 to specify a percentile.
- RiskKumaraswamy(alpha1,alpha2,minimum,maximum) specifies a Kumaraswamy distribution with the defined minimum and maximum and shape parameters alpha1 and alpha2. The Kumaraswamy is a very flexible distribution that can be used as a mathematically simpler replacement for the BetaGeneral distribution.
- RiskReciprocal(minimum,maximum) specifies a Reciprocal distribution with the defined minimum and maximum. The Reciprocal distribution is commonly used in numerical analysis, Bayesian statistics, and analysis of noise.
Other alternatives to some of the functions above are the corresponding "AltD" functions where any entered percentile values are cumulative descending percentiles, where the percentile specifies the probability of a value greater than or equal to the entered value.
New Statistical Functions
Statistics functions return desired statistics on simulation results for 1) a specified cell or 2) a simulation output or input. These functions are updated in real time as a simulation is running or at the end of a simulation. Statistics functions located in template sheets used for creating custom reports are updated only when a simulation is completed. All of the following functions can use the RiskTruncate property function to optionally restrict the range of the simulated distribution for calculating the statistic.
- RiskCIMean(cellref or output/input name,confidence level,lower bound,Sim#) returns the lower or upper bound of the confidence interval of the mean of the simulated distribution for cellref. Instead of a single point estimate for the mean, a confidence interval generates lower and upper bounds for the possible value of the mean, at a given confidence level.
- RiskCoeffOfVariation(cellref or output/input name,Sim#) returns the coefficient of variation of the simulated distribution for cellref. The coefficient of variation is a measure of dispersion of a probability distribution or frequency distribution. It is often expressed as a percentage, and is defined as the ratio of the standard deviation to the mean.
- RiskMeanAbsDev (cellref or output/input name,Sim#) returns the mean absolute deviation of the simulated distribution for cellref. Mean Absolute Deviation is the mean of the data's absolute deviations around the data's mean or the average (absolute) distance from the mean.
- RiskSemiStdDev (cellref or output/input name,lower_data,Sim#) returns the semi standard deviation of the simulated distribution for cellref, or the standard deviation of the values in the distribution below the mean.
- RiskSemiVariance (cellref or output/input name, lower_data, Sim#) returns the semi variance of the simulated distribution for cellref, or the variance of the values in the distribution below the mean.
- RiskStdErrOfMean (cellref or output/input name,Sim#) returns the standard error of the mean of the simulated distribution for cellref.
Run Goal Seek, Evolver or Solver During Each Iteration without VBA Coding
For EACH iteration in your simulation you can now run Goal Seek, Evolver or Solver without having to face the hurdle of writing a VBA macro to do so. You can simply set this up using the Macros tab of the Simulation Settings dialog, and @RISK does the rest for you - saving you an incredible amount of time and hassle!
Graphing and Reporting Improvements, Support for Excel Themes
All @RISK graphs and reports now feature a more streamlined interface, to accomplish more in fewer clicks. In addition, graphs and reports support Excel’s built-in themes and colors. Now you can apply your company’s standard Excel themes to your @RISK reports, improving communication consistency and efficiency.
New Statistics Functions Tab
Easier Access to Default Graphs
Greater Control over CPU Usage
Apply to Open Models Checkbox in Application Settings Dialog
Optimized for Windows 10 and Excel 2016
All DecisionTools Suite products, including @RISK, have been retooled for optimal performance and presentation in the latest Windows 10 and Office 2016 environments.
The user interface has been updated to make common tasks easier to perform, saving time and clicks.
Enhanced Support for Ultra HD Displays
We’ve leveraged the power of ultra high definition displays to ensure that all DecisionTools graphs and reports look better than ever.
Ability to switch between Languages More Easily
Switch between any of the language versions without having to change a corresponding code page.
Other Changes of Note
In addition to @RISK's New Features the following Maintenance Fixes have also been taken care of.
*** A Selection of Significant Maintenance Fixes ***
12541 “Swap Out” Summary Report sometimes displays incorrect thumbnail graph image.
12475 'Overflow' message when running Stress Analysis with very large number of iterations.
12447 Swap Out with protected sheet doesn't produce thumbnail graphs.
12373 @RISK for Project resource sheet items improperly hidden in international versions.
12366 RiskCurrentSim returns incorrect value during “After Simulation” macro.
12360 Output marked with RiskIsDate property has incorrect number formatting in Excel reports.
12358 Multiple RiskProjectRemoveTask function might interfere with each other.
12353 @RISK for Project Gantt chart incorrectly suppressed.
12348 @RISK for Project Import.MPP feature not working correctly with Project 2007.
12346 Probabilistic calendars dialog does not appear for some models.
12344 Corrected correlation matrix still not self-consistent.
12339: Theoretical statistic functions applied to RiskMakeInput function returning incorrect values.
12333 @RISK for Project Parameter Entry Table not adding fields all selected tasks.
12324 Smart Sensitivity Analysis failing to recognize subtasks.
12319 @RISK for Project constraints being overridden.
12316 Accessing tornado settings via Excel right-click menu can freeze @RISK.
12314 Color cells and thumbnail options may slow down Excel significantly in higher DPI display modes.
12310 Infrequent crashes when saving simulation data while color cells options on.