ModelRisk release notes

The Turbo engine now supports more Excel and ModelRisk functions, and its results match those of the classic simulation engine on a wider range of models. Overall stability has also been improved.

Minor fixes in the Optimizer window.

A new Optimizer has been added to the ModelRisk ribbon (ModelRisk Optimizer). It searches for the values of one or more worksheet cells — the decision variables, shown in the Inputs grid — that give the best value of one or more output statistics — the objectives, shown in the Targets grid, where several objectives are combined into one score — while respecting any constraints you set. Because the model may contain Vose distributions, each candidate is evaluated by a Monte Carlo simulation, so the Optimizer works directly on uncertain models.

Key capabilities:

1. Objectives (Targets grid). Pick the output cell's Statistic — Value, Mean, Median, Mode, Min, Max, StdDev, Percentile, Probability (above, below or between values), CVaR (lower or upper tail), SixSigma and more — and a Mode: Maximize, Minimize or Equal to Goal.
2. Decision variables (Inputs grid) of several Types: Continuous, Integer, Discrete (List), Discrete (Step) and Binary. Min / Max bounds and the Step can be numbers or cell references; a Binary variable is 0 or 1, and a Discrete (List) one takes its values from a list.
3. Constraints. Regular constraints compare a statistic of an output cell with a bound (<=, <, =, >, >= or between); Formula constraints point at a worksheet cell that must be TRUE. Each is set as Hard or Soft, where a Soft constraint applies a penalty with a weight.
4. Groups that optimise several variables together under a method: Independent, Budget, Sequence or Assignment.
5. Method (engine). Auto chooses the engine from the model's structure, or you can force Vertex Detector, Projected Gradient, BFGS, Brent, Genetic or Goal Seek.
6. Type. Stochastic runs a full Monte Carlo simulation per trial; Deterministic recalculates the model once per trial, for models without uncertainty.
7. Faster evaluation. Use Turbo Simulation runs the whole optimisation on the Turbo engine (see below). On the Advanced Settings tab, the Evaluation option Prefer Central Values runs each trial at the median of every distribution — available when all objectives and constraints use Mean or Value — and Detect Deterministic Models, on by default, switches a model with no live uncertainty to one recalculation per trial.
8. Stopping controls. Tolerance, Stagnation, maximum Trials / Time, Stop on error and a custom Stopping formula; VBA macros at the start and the end of the optimisation and before / after each trial; and a Seed for reproducibility.
9. Results. An Overview of the best solution, Solving History (convergence chart, trial journal and reports), Apply Solution to write the values back to the sheet, and Efficient Frontier Analysis (Start Start Efficient Frontier Analysis).
10. Scenarios. Several optimisation models per workbook, saved with the workbook.

Depending on the structure of the model, Turbo Simulation runs from several times to several tens of times faster than a classic simulation. The new engine computes the model outside Excel; the results go to the workbook and the Results Viewer as usual.

1. Ribbon. The Start drop-down (ModelRisk Simulation) has a new item, Start Turbo Simulation. The main Start button runs Turbo when Use Turbo Engine by Default is set.
2. Simulation Settings, Model Settings tab. Use Turbo Engine by Default (for this workbook) and Turbo Consistency — before the first Turbo run of a workbook ModelRisk offers to check that Turbo reproduces the classic engine on it. The status (Verified / Not verified) is kept in the workbook and can be cleared with the button next to it.
3. Simulation Settings, Application Settings tab. Use Turbo Engine by Default for New Models and Turbo Consistency — the defaults for workbooks without their own settings.

Distributions in the Extreme Scenario Analysis window can now be combined into user-defined groups. A group is a unit of the analysis in its own right, exactly like an individual distribution: when its impact is assessed, the extreme constraints are applied to every distribution in the group at once, the output is read, and all members are then restored — so you can measure the combined effect of several inputs reaching their extremes together.

1. Add Group, next to Add Input distributions, creates a group and opens the Group Settings window; Group Settings can also be reopened for the selected group from the ribbon.
2. Group Settings holds the group Name, its Scenario options (edited once and applied to the whole group), and the group Members, managed with the Add / Remove / Clear List links. Add Group Members lists the available distributions with checkboxes, all selected by default.
3. The input list gains a top grouping level: Independent Distributions (each distribution simulated on its own, as before) and every user group, with the worksheet as a nested level underneath. The Independent and group header rows are colour-tinted for clarity.
4. Inside a group the members' Include and Scenario options are shown greyed and read-only, because participation and the extreme parameters are controlled by the group.
5. A distribution can belong to several groups and still remain in Independent Distributions. Each membership yields its own result series, so groups appear in the reports and charts on an equal footing with ordinary distributions.
6. A new Name column (after Distribution). For Independent distributions the name can be edited in place, and the change is reflected everywhere the distribution appears, including every group it belongs to.

VoseSimMode(Cell Reference, Simulation number) returns the mode — the most likely value — of the referenced cell over the simulation, in line with the existing VoseSimMean, VoseSimStdDev and similar functions. The optional Simulation number selects which simulation to report on; it equals 1 if omitted.

The Network Monitor has been improved to provide more reliable network status monitoring and diagnostics. Further refinements enhance stability and overall performance.

1. Truncation bounds given as cell references (RiskTruncate, RiskTruncateP) are now kept instead of being replaced or dropped.
2. Formulas using RiskName inside a larger calculation keep the surrounding calculation, for example a multiplier.
3. RiskPertAlt and RiskTriangAlt with "min", "m. likely" or "max" now convert correctly instead of returning #VALUE!.
4. Cells no longer become array formulas by mistake, which also fixes the "Unknown error converting this formula" message.
5. RiskMode now converts to VoseSimMode.

1. Silent installations can now set the Excel version, e.g. ModelRiskSetup.exe /quiet OfficeBitness=x64.
2. Fixed shortcut warnings that could interrupt the installation.
3. Improved detection of the Excel version (32-bit or 64-bit) during installation.

Corrected the behaviour of the VoseCopulaMultiNormal and VoseCopulaMultiT copulas in multi-simulation mode.

The Goal Seek tool can now drive an input cell until the cumulative probability of the simulated output (Target) reaches a value you specify. Two new metrics are available in the Metric list:

1. Probability( Target > Value ) — Goal Seek adjusts the input so that the probability of the Target output exceeding the given Value equals the Target Probability you set (90% by default).
2. Probability( Target <= Value ) — the complementary metric: Goal Seek adjusts the input so that the probability of the Target output being less than or equal to the given Value equals the Target Probability.

When one of these metrics is selected, the target-value field is renamed to "Target Probability" (the probability you want to reach, e.g. 90%), and an extra "Value" field appears for the threshold against which the probability is measured.

Fixed an issue that could occur in certain workbooks with a specific internal structure. In these cases, the analysis could not be started because the Start Analysis button did not respond to clicks.

Fixed an issue where, in certain workbooks, clicking the button did not enable highlighting.

A new "Show Deviations" check box has been added to the Tornado Options ribbon. When it is enabled, the value labels at the two ends of each tornado bar show the deviation from the baseline (bar-end value minus baseline) instead of the absolute conditional values, making it easy to read how far each end of a bar moves away from the baseline in either direction. When the option is off, the bars display the usual absolute minimum/maximum values. The setting is stored with the chart and restored when the workbook is reopened.

A new editable "Baseline" combo box has been added to the Tornado Options ribbon. By default it is set to "Auto", which places the baseline (the dashed reference line) at the computed median of the output. You can now type any numeric value to pin the baseline at a fixed position; selecting "Auto" again returns to the automatic median. When a custom baseline is set, the chart's horizontal axis is automatically rescaled and centred on the baseline so that the line always stays visible — even when the value lies outside the range of the bars — while keeping every bar within view. The baseline value also drives the deviation labels described above. The setting is stored with the chart and restored when the workbook is reopened.

A new "Auto Selection" toggle button has been added to the Tornado Options ribbon (Manage Inputs panel) next to "Filter to Precedents". When enabled, the Tornado chart automatically keeps only the inputs that are precedents of the currently selected output — the input list is rebuilt on every output switch and refreshed after each new simulation, with no manual filtering needed. While the toggle is ON, the input checkboxes in the left panel are locked (a notice explains how to unlock them); turning the toggle OFF re-enables manual editing and leaves the current chart untouched. Precedent scans are cached per workbook on disk and reused across output switches and RV restarts; the cache auto-invalidates when the simulation session changes.

Fixed a bug where the Tornado chart could appear completely blank when at least one selected input had an extremely long name (>35 characters). The long name was being used in full as a Y-axis label and squeezed the plot area down to zero width. Variable names on the Y axis are now truncated to 35 characters with an ellipsis ("...") when longer; the full name is still shown in the bar tooltip.

A new "Filter to Precedents" button has been added to the Tornado Options ribbon. Clicking it scans the active workbook for cells the currently selected output formula actually depends on, then opens a dialog listing every Tornado input that is NOT a precedent of that output. On Apply those inputs are removed from the chart — only the meaningful contributors remain.

1. Scan Inputs dialog. The "Exclude from results" group with 8 advanced exclusion checkboxes has been moved out of the main dialog into a new "Advanced..." sub-dialog opened by a button. The main Scan Inputs dialog is now compact and less intimidating; defaults are unchanged.
2. Scan Results dialog. The "Tier" column has been removed; only the "Score" column remains. Rows with Score ≥ 0.70 are still pre-selected automatically. Column widths have been redistributed so they fill the full grid width.
3. The "Impact" tab is now hidden until the analysis has been run for the first time, so the user cannot accidentally switch to an empty results page.
4. Tornado chart on the Excel report sheet. Each Tornado bar is now drawn in two colours: the segment from Min to Expected Value in dull red, the segment from Expected Value to Max in the current theme bar colour. This matches the in-dialog Tornado view.

Default colour palette refreshed: Inputs = light green, Outputs = light red, Top Output Drivers = light yellow, all other categories = light blue. Inputs / Outputs / Top Output Drivers are bold by default.

VoseSimTVARp(X, p) returns the mean of the simulated values that fall below the probability level p of the CDF. VoseSimTVARx(X, x) returns the mean of the simulated values that fall below the threshold value x.

Both functions reuse the existing VoseSimCVARp and VoseSimCVARx algorithms applied to the negated input, exploiting the symmetry between the two tail conditional expectations:

This guarantees full numerical consistency with the existing CVaR statistics and requires no changes to the underlying simulation engine.

All TimeSeries functions — VoseTime, VoseTimeFit, VoseTimeMulti, VoseTimeMultiFit and their variants — can now be inserted into a single cell. In this mode the new "Forecast Periods" parameter defines the range over which the function spills, taking full advantage of Excel 365 dynamic arrays — no need to pre-select a range or array-enter.

Deep optimisation of the VoseTimeMulti family delivers a 10–20x speed-up in simulation, with even larger gains on bigger models. The improvement comes from an intelligent caching engine for Cholesky factorisations of large covariance matrices, which removes the dominant per-timestep cost of multivariate time series simulation. Model accuracy and function signatures are unchanged.

1. The VoseName property is now visualised inside the Bivariate Copula and Multivariate Copula windows: custom names attached to component distributions are picked up and displayed on the corresponding charts and controls.
2. Sparkline colours automatically adapt to the chosen Highlight background, keeping in-cell sparklines readable under any colour setting.
3. Fixed a configuration-dependent issue that prevented the fx(x) and F(x) windows from opening on certain user setups.

Instantly visualise the structure of your model: ModelRisk highlights every Input, Output, Distribution, Copula, Fit function, Time Series and Risk Event with a fully configurable colour scheme.

Find the input value required to hit a target output metric — mean, standard deviation, variance, skewness, kurtosis or any percentile — through automated iterative simulation. This release removes a key limitation: Goal Seek now fully supports models that use conditional logic inside cells when running in Fixed Seed / Goto Sample mode, thanks to a redesigned RNG state-preservation algorithm.

Stress-test your model under any combination of scenarios and input distributions. Apply user-defined scenarios to multiple inputs at once, simulate, and compare output behaviour under extreme conditions via histograms, CDF curves and box plots. CDF axis scaling has been corrected, and the reporting engine now renders correctly across all locales.

A dedicated tool for identifying the inputs that most influence your model. Define deviation ranges for any set of inputs — Uniform, Triangle, Beta, PERT or any custom distribution — run a single sweep, and instantly see which factors drive your outputs through tornado charts and detailed impact rankings. Fully compatible across all language versions.

The simulation engine has been optimised, enabling multi-fold speedups for certain models. The actual performance gain varies and depends on each model's structure and configuration.

Fixed an issue in VoseSimProbability that occurred when the number of simulations was ≥ 2 and idSimulation was set to a value less than the total simulation count. In this case the function could incorrectly return 0.

Added an option to hide the "Save simulation results" prompt, which appears when running a simulation again.

Previous versions of the SimSensitivityMean and SimSensitivityPercentile functions did not support operation when optimisation was disabled. Also fixed an issue in parsing the model's currency settings that could unexpectedly cause the Simulation Results view window to close.

VoseOutput and VoseInput functions, written using the "@" symbol and a space, were not removed.

Clicking Remove Output/Input with multiple cells selected in a large workbook caused the workbook to recalculate multiple times.

While using the "_" character in a custom cell format together with a currency symbol (intended to hide the currency symbol), the currency symbol was still displayed in charts.

An error occurred when adding values close to the upper limit for floating-point numbers to the graph.

1. A new window for selecting correlated distributions has been added. It now pops up immediately before the correlation settings window opens.
2. The "Correlation" menu has been replaced with a split button labelled "Correlation". When the button is clicked, the correlation type (multivariate or bivariate) is determined automatically based on the number of selected distributions.
3. The dropdown menu with correlation types remains available and allows the user to select the desired type manually. In this case no preliminary analysis of the number of distributions is performed, and the window for the selected correlation type opens directly.

Chart tabs with names including specific character combinations worked inconsistently.

Unexpected Excel shutdown during the simulation with optimisation options disabled.

During fitting parameters for GBM, selecting an array of historical time stamps from the worksheet resulted in an error message.

In response to recent Microsoft Office security hardening measures, we have implemented a revised approach for accessing and managing satellite files within the Excel add-in environment.

1. A new Application Settings window allows default settings to be set for new charts. Export and Import operations are also available, so settings stay consistent within an organisation.
2. A "Use active chart settings as default Application settings" tool has been added. It applies the selected chart settings as the defaults for new charts.
3. Support for displaying symbols other than dollars in results has been added for currency variables.

1. A new "Variables" list displays the variables within each of the SIDs in the library. A mini-histogram and statistical coefficients are displayed for each variable.
2. A new "Data Viewer" menu allows the selected SIDs or individual variables to be viewed on charts.
3. A new "Variables" menu (activated when the user switches to the "Variables" list) allows new SIDs to be created from the variables of existing files, and selected variables to be deleted from a SID.

1. The acceptable range of the "Alpha" parameter of VoseCopulaMultiClayton and VoseCopulaBiClayton is extended from 0 to 150.
2. The acceptable range of the "Theta" parameter of VoseCopulaMultiGumbel and VoseCopulaBiGumbel is extended from 1 to 10.
3. The acceptable range of the "Theta" parameter of VoseCopulaMultiFrank and VoseCopulaBiFrank is extended from 0 to 670.

Introducing the new optional parameters MinLimit, MaxLimit and DistrShift, which can be applied to the severity distribution before aggregation.

Support for exporting results of VoseSim functions.

1. Support of RiskCopula functions for @RISK to ModelRisk conversion.
2. Support for the RiskCompound function conversion when parameters are defined as distributions.

1. VoseTriangleAlt6 — returns a random value from a Triangle distribution specified by three percentiles.
2. VoseGammaAlt — returns a random value from a Gamma distribution specified by three percentiles.

1. Model opening bug: the ModelRisk icon closed unexpectedly when opening a document from an untrusted location.
2. SIDs Library bug: an attempt to delete a SID located in a subfolder of the current folder using the context menu resulted in a crash.
3. The video link in the ModelRisk menu bar now goes to the Vose Software YouTube channel.
4. The tool for selecting custom program folders now allows preferred working directories to be specified (ModelRisk icon "Advanced settings" "Change ModelRisk directories").

1. Quick edit of slider or marker position by clicking on it.
2. Added decimals control to the marker settings page.

1. New design for chart markers and sliders — the appearance of charts is improved.
2. New customisation tools for markers and sliders: custom colours can be set for the background and text of labels, or these parameters can be bound to the colour of a variable.
3. A new data export tool for the "Statistics" chart — "Export statistics". Available on the "Statistics options" tab in the "Export to XML" menu, it exports statistical coefficients and percentiles (a custom step for percentiles can be set).
4. One taskbar tab instead of several (the old version used one tab per chart).
5. Optimisation of memory use when exporting large amounts of data to XML.

A new "University" licence has been created — a type of network licence that grants permission to use the software within a local network. Activation and connection are identical to the network licence, but it has two advantages:

1. The allowed number of concurrent users is unlimited.
2. After a successful connection to the university licence, the client workstation can continue to operate autonomously for the validity period of the licence — teaching staff do not need separate personal licences, and students can use the products for self-study.

Automatic recovery of deleted network / university licences has also been implemented.

A simulation results histogram sparkline is displayed for each cell containing the VoseInput / VoseOutput function. This can be turned off for a specific model using the "Show results sparklines" switch on the "Model settings" tab of the Simulation Settings window, or application-wide on the "Application settings" tab.

1. Select the cells containing ModelRisk distributions (hold Ctrl to select non-consecutive cells).
2. Right-click and select "ModelRisk correlate". Depending on how many cells were selected, the Bivariate Copula or Multivariate Copula window opens with the selected distributions loaded.
3. Set the required correlation coefficient and click "Insert to worksheet" to complete the setup.

1. Export and import of data in SIPMath® Standard 3.0 (probabilitymanagement.org/sipmath) and CSV formats.
2. Support for switching between the full and relative path to a SID file.
3. Support for using the path to a CSV or SIPMath® Standard 3.0 file as a data source for the VoseSID function, without converting it to SID format.
4. Support for viewing SIPMath® and CSV files.
5. Support for switching between several SID Library locations.
6. Support for opening the SID Library in Windows Explorer.

1. Support for changing the number of decimal points in the Statistics pane via the "Statistics settings" button at the top of the pane.
2. Support for changing the number of decimal points shown on chart sliders in the "All options" interface.
3. Additional tools in the Correlation plot. The Joint success delineation tool splits the scatter datasets into areas below and above the target joint probability defined in the "Probability" field, and lets you set whether more is better or worse for each variable. The Target Point tool splits the data into four areas and shows the fraction of data points in each — useful for questions such as how likely an NPV above 30M is, given market growth above 30%. The Median Point tool adds a large dot showing where the median of the joint distribution lies.
4. A new Boxplot view: the "Whiskers-plot".

Added an "Export results" button to the "Results" menu in the ribbon, which saves the simulation output data in the following formats:

1. SID format. Once exported, a SID with results is placed in the SID Library and can be used in other models via the VoseSID function.
2. SIPMath® Standard 3.0 file.
3. XML file.

A new visual appearance template arranges the ribbon elements as close as possible to the @RISK menu. It helps those migrating from @RISK and is available via "Help" "Select appearance" "Standard".

A tool to view the list of ModelRisk functions in the current model (it replaces the "Find" tool of older versions). Apart from easy navigation through the ModelRisk functions in a model, it offers:

1. Bulk add/remove of Inputs and Outputs on model cells.
2. Switching between showing values and cell references in cells with ModelRisk functions.
3. Bulk-set cell appearance for cells with ModelRisk functions.
4. A report on the ModelRisk functions used in a model.

Allows Picture-Shape objects to be created in a spreadsheet model that are updated after simulation runs, based on Results Viewer graphs.

VoseStatic returns a fixed value for the distribution when random number generation is disabled.

1. New conversion in the ModelRisk Converter: RiskStatic (@RISK) to VoseStatic (ModelRisk).
2. Static non-volatile versions of four distributions added: LognormalAlt3U, LognormalAlt6U, LogLogisticAlt6U, WeibullAlt3U.

1. Rendering bug in the Multivariate Copula and Bivariate Copula windows, reproduced when the windows were resized.
2. The "Microsoft Excel is waiting for another application to complete an OLE action" message was shown on Excel / ModelRisk startup.
3. "SWAP" tool stability in localised versions of MS Office.

The tool creates a copy of a ModelRisk model where all ModelRisk functions have been removed and replaced with fixed values. Functions that generate random samples are replaced with their median value. This allows a working version of the model to be shared with non-ModelRisk users, for example for external auditing. Optionally, a note is automatically attached to any edited cells showing the original formula.

Primarily for @RISK users migrating to ModelRisk; this also extends the ability of ModelRisk to convert @RISK models automatically.

1. Redesigned LognormalAlt3 — sets a Lognormal distribution using three alternative parameters: the mean, a random value and the cumulative probability of that value.
2. WeibullAlt3 — sets a Weibull distribution using the mean, a random value and the cumulative probability of that value.
3. LognormalAlt6 — sets a Lognormal distribution using three percentiles, each given as a value and the probability that other values are lower. Percentiles can be set freely.
4. LogLogisticAlt6 — sets a LogLogistic distribution using three percentiles in the same way.

ModelRisk now converts over 98% of the functions used by the other products.

1. RiskJointProb2Dim VoseSimMean
2. RiskJointProbMultiDim VoseSimMean
3. RiskNormalAlt and RiskNormalAltD VoseNormalAlt
4. RiskLognormAlt and RiskLognormAltD VoseLognormalAlt6

1. An additional optional parameter VariableIndex has been added to the VoseSID function, which returns values from only one of the variables in the SID file. The function can also now be placed in a single cell (previously an array), such as in an Excel Data Table.
2. A new function VoseGetVarName(NameSID, VariableIndex) returns the variable name from the specified SID file, making it easy to identify the right variable when a SID file contains several.
3. A window for additional parameters of the SID function appears when the VoseSID function is placed from the SID Library into a worksheet. It configures the value sampling mode (sequential iteration, random sampling, or a value from a fixed row) and the number of variables used (one or all).

VoseCVARp and VoseCVARx return the CVaR from an array of simulated data located within the spreadsheet.

1. Tornado charts in the Results Viewer in "Conditional Cumulative Percentile" mode did not react to a manual change of the percentile value.
2. Behaviour when exiting Excel has changed: by default ModelRisk now closes completely when the user exits Excel. The previous mode can be restored in the Options interface.

1. New "Create SID" button on the "General" tab — uploads all simulation results, or the active chart data only, to the SIDs Library.
2. New "Export to XML" button on the "General" tab — exports all simulation results or the active chart data only.
3. New "Export to XML" button on the "List options" tab — exports the data list.

1. The toolbar has been redesigned.
2. New "Create SID from Simulation results" imports the active model simulation results, or saved results from .vmrs files, into the SIDs Library.

1. Correlation values have been added to the "View data" tab, for quick comparison of the rows or columns selected for viewing.
2. The correlation matrix on the "Multivariate Data Analysis" and "Multivariate time series" tabs has been redesigned.
3. Several changes in the other Data Analysis windows improve the visibility of the data presentation.

The "Optimisation options" window has been redesigned.

1. The simulation performance of large models has been optimised.
2. A new high-performance algorithm for model analysis, used in the initial stage of simulation, has been implemented.
3. The performance of the simulation statistics functions has been improved, in some cases significantly: VoseSimPercentile, VoseSimProbability, VoseSimCorrelation, VoseSimCorrelationMatrix, VoseSimCofV, VoseSimCVARp, VoseSimCVARx, VoseSimSensitivityMean, VoseSimSensitivityPercentile, VosePrecisionProbability, VosePrecisionPercentile and the VoseSixSigma family.

1. VoseSID: performance improved, compatibility bugs with the new algorithm fixed (6.1.91, 6.1.92, 6.1.93), and incorrect behaviour in Fixed Seed and Goto Sample modes fixed.
2. VoseCopulaMultiNormal and VoseCopulaMultiT performance.
3. Results Viewer: when selecting the list of results from the first to the last line manually and copying, the last line was not copied to the clipboard.

1. Results Viewer: new interface design of the main controls, a new Statistics pane supporting percentile sorting, and the read-only flag after saving a .vmrs file has been removed.
2. New interface design of the Select Distribution window, including a new distributions grid and a new statistics table.
3. The new statistics table is implemented in all windows where it is supported.

1. Weibull mathematics used for the known x1/p1, x2/p2 was incorrect for "beta".
2. Exporting a Distribution Object failed.

1. Results Viewer Data List tooltip text overflow.
2. Slider bug in some windows: an error occurred when moving chart sliders on non-English systems.
3. Expert window drawing positioning bug, which occurred while the window was in full screen mode.
4. Expert window localisation bug: in non-English localised versions of MS Office the recalculate button was unresponsive.
5. Aggregate Panjer window: selecting any distribution in the Fit menu could cause an application crash.
6. Aggregate DePril window slider bug, which occurred when the marker values of the Statistics grid "Lower P" or "Upper P" were entered manually.
7. Results Viewer: the initial position of the slider in the CumulA and CumulD charts could be incorrect when the minimum value of the simulation results repeated in more than 5% of all values.

VoseCopulaMultiNormal and VoseCopulaMultiT performance has been optimised. The optimisation settings are managed through the Advanced Options item of the ModelRisk taskbar icon context menu, under the Optimisation Options button.

1. Copula insertion into linked distributions combined with VoseInput or VoseOutput (Bivariate and Multivariate Copula windows).
2. Prevented creation of the hidden DSNFiles sheet when there are no associated DSN files.
3. Results Viewer: incorrect reading of the Type option on the Variables tab of the Histogram chart settings (an undefined Type was saved as Line instead of Bar).
4. Results Viewer: incorrect slider value formatting on non-English locales.
5. Pasting formulas in uppercase, reproduced in some localised versions of Office, for example Turkish.
6. User Interface Template loading error: the NUMLOCK key automatically turned off when starting ModelRisk and switching the ribbon template.

The Advanced Options window, available from the context menu of the ModelRisk taskbar icon, has a new Optimisation settings button that manages Simulation Statistics Optimisation and Distributions Optimisation at simulation time.

VoseSimPercentile, VoseSimProbability, VoseSimCVARp, VoseSimCVARx, VoseSimCorrelation, VoseSimCorrelationMatrix, VoseSimMean, VoseSimVariance, VoseSimStdev, VoseSimCofV, VoseSimSkewness, VoseSimKurtosis, VoseSimMin, VoseSimMax, VoseSimSemiVariance, VoseSimSemiStdev, VoseSimMSE, VoseSimValue and VoseSimTable.

1. VoseSimPercentile: a problem occurred when the workbook name included a space.
2. Results Viewer: removed tooltip flickering when moving the mouse cursor over the list of variables.
3. Cholesky algorithm: a problem in calculations when the input matrix consisted of zeros and ones.
4. Pasting formulas in uppercase, reproduced in some localised versions of Office, for example Turkish.
5. Goto Sample simulation mode.
6. Advanced Options window: during initial loading, a "Server busy" system message sometimes appeared.

Fixed the behaviour of the Distribution Object functions when simulating in the mode with active optimisation.

Fixed the functioning of VoseProb and VoseDataObject, and the interface for creating them.

1. The Select Interface Templates demo window.
2. Improved simulation algorithm in Fixed Seed mode.
3. Improved Results Viewer design.

Support for the automatic connection of workstations to the network licence.

Fixed a bug reproduced in Stop On Output Error mode when simulating workbooks without Vose functions.

Fixed an activation bug in 2-year and 3-year licences.

1. A new distribution, VoseThreePercentiles.
2. French language added.
3. An overall License Manager manual for our products.

1. Improved parallelisation algorithms for some complex features that use a cache.
2. Updated interfaces for the non-English languages.

1. AutoRun on user logon did not work when switched off and on again.
2. A problem that caused the "ModelRiskSettings.ini not found" message.