Common elements of ModelRisk windows | Vose Software

Common elements of ModelRisk windows

See also: ModelRisk functions and windows, Statistical descriptions of model outputs, Graphical descriptions of model outputs, Graphics, workflow and error handling in ModelRisk

Certain elements appear in many ModelRisk windows. For example, a table with summary statistics; graphs; and fields like the output field.

This topic gives a more in-depth explanation about these elements and how to use them.

Fields and buttons - Output location, M/C, Generate, Help, OK, Cancel

Most of the ModelRisk windows correspond to a VoseFunction (e.g. an aggregate distribution, sampled value from a spliced distribution, etc...) being inserted in one or more spreadsheet cells. The location where the VoseFunction is to be inserted in the spreadsheet, is specified in the Output Location field. The desired VoseFunction will be inserted upon pressing the OK button, after which the ModelRisk window closes.

Pressing the Cancel button closes the window without modifying the spreadsheet in any way. Pressing the Esc keyboard button has the same effect.

The Generate button re-generates the random values for the preview graph(s) in the window.

With the M and C buttons you can switch between viewing the Probability Density (or Mass), or cumulative graphs of the distribution(s) shown.

Summary statistics table

Distributions and their properties play a central role in risk analysis modeling.

In all ModelRisk windows that have a function related to distributions (which many do), a table with summary statistics of the currently selected/loaded/previewed distribution is shown on the right. For ease of use, this table always contains the same elements, regardless of the specific ModelRisk window it is in. These elements are:

Location
- Mean
- Median
- Mode
- Min
- Max

Spread
- Standard deviation (St. dev.)
- Variance
- Coefficient of Variation (CofV)

Shape
- Skewness (S)
- Kurtosis (K)

Where it is relevant, the following statistics are shown as well:

Markers

Markers can be edited in the statistics pane. Use markers to see the x-values corresponding to percentiles, and vice versa. For example, the image shows what you see when setting the LowerX value to -1 and the UpperX  value to 1 for a normal(0,1):


- LowerX - the lower marker value (-infinity by default)
- UpperX - the Upper marker value (+infinity by default)
- LowerP - the Lower marker's  Percentile value (0 by default)
- UpperP - the Upper marker's Percentile value (1 by default)
- X Spread - the range between the markers, i.e. UpperX - LowerX
- P Spread - the percentile range between the markers, i.e. UpperP - LowerP

The marker settings shown in the image correspond to this PDF plot:

Percentiles
The 1st, 5th, 25th, 50th, 75th, 95th and 99th percentiles are always shown. If the ModelRisk window is enlarged (by dragging its lower right corner), additional percentiles are shown.

Preview graphs

In windows where distributions play a role, the PDF or PMF, CDF or both graphs are shown.

When you hold the mouse pointer above a graph, it comes "in focus" and all other visible elements are greyed out, for easily pointing somebody to a certain graph.

The graphs update dynamically according to the distribution's parameter values when these can be specified within the window.

On top of graphs are the graph toolbar buttons. Below is explained what you can do with the most important of these buttons. Note that these buttons are not necessarily all present in a given ModelRisk window.

  • Copy to clipboard, either as a bitmap, metafile or text (data only). This allows you to use the graph image in other programs, like MS PowerPoint, or for your web site... (try copying as bitmap, and then pasting in MS Paint by pressing CTRL+V)
    - To copy the image as you see it presented, including colors etc... choose As a bitmap. This is the "normal" way to copy an image to the clipboard, like you are probably used to doing in Windows.

    - You can also copy the image to the clipboard as a metafile (WMF), the Windows vector file format. An advantage of vector graphics is that they can be rescaled without quality loss (in the proper program). Most fonts in Windows are basically vector graphics, for example.

    - You can also copy the (randomly sampled) source data, used to construct the preview graphs, as text to the clipboard. For a PDF graph this is a set of (x,y) pairs, for example.


Export options
in the Empirical Copula window
  • Print. This will take you to the common Windows dialog for selecting a printer, adjusting printing options, and of course print the graph(s).

  • Palette Selector. You can adjust the graphs colors to one of the preset color palettes. Note that every graph element can be formatted individually as well by right-clicking it.

  • Zoom. To zoom in to a part of the plot, press the Zoom button and then drag a rectangle in the graph area. This will zoom in to this graph. To zoom out again, press the Zoom button again.

  • Edit X axis. This allows you to set the boundaries shown for the X-axis, or to toggle "auto scale mode" (to scale the X axis automatically).

  • Properties. Change more properties of the graph, like colors of every element, font formatting, 3D view...

  • Export options. Whenever a ModelRisk window has more than one type of output to insert in the spreadsheet, these Window-specific types of output can be accessed with the Export options button above the preview graph.

Industrial version only

  • Create report. Clicking this button will produce a fit report in a new Worksheet with the fitted models in a table. The table will have the fitted objects, rankings, statistics and percentiles of the fitted models. (This depends on object type) The report will also include the OptimatFit function that automatically returns the best fitted model according to the selected information criteria.

 

 

Whenever a preview graph has boundaries (visualized by vertical lines), these can be dragged along the graph to change the corresponding boundary value.

By right-clicking anywhere in the graph area, you can quickly access many of the above through the context menu that pops up.

Errors

In ModelRisk windows, a descriptive error message is shown in red when appropriate. For example, the bivariate copula requires an Alpha parameter between -35 and 36. If the currently chosen Alpha lies outside of this range, this will be pointed out in red. For more information about errors in ModelRisk, click here.

For general troubleshooting, also see the FAQ - Troubleshooting topic.

 

 

Navigation