Options

• 7/23/2022
• 4 minutes to read

This topic describes the various options that are available in Bayes Server.

Show Start Page

When True, the Start Page is shown when Bayes Server is started.

Migrate Existing Distributions

When structural changes, such as adding or deleting nodes, links and variables, are made to a Bayesian network or Dynamic Bayesian network, certain distributions become invalid as a result of the changes.

• When False, invalidated distributions are set to null.
• When True, invalidated distributions are amended, so that as much information as possible is retained.

Node toolbars

This option determines whether a Node Toolbar is displayed on each node.

Default state names

Defines the default states added to a new discrete node.

Maximum discrete chart items

Defines the maximum number of discrete distribution combinations that should be charted in Query explorer.

If a distribution has more combinations than this, the values of the distribution can still be viewed by double clicking on the query in Query explorer.

Requery on add

Determines whether a re-query should automatically occur, when new distributions are added to Query explorer.

Standard deviations

Defines the number of Standard Deviations to use when plotting Gaussian distributions.

Merge Gaussian mixtures

When True, plots of univariate mixtures of Gaussians use a combined weighted distribution. When False, each Gaussian mixture is plotted separately.

Maximum Undo/Redo items

Defines the maximum number of items to be held in the Undo/Redo list per network.

Initialize new distributions

Determines whether distributions are automatically created and initialized, when new nodes are created, or new temporal links are created that require new distributions.

Discrete distributions are initialized to a uniform distribution.

Continuous (Gaussian) distributions are initialized to have zero mean, and variance of 1.

Case Column Names

When the Data map window is opened, variables are automatically mapped to data columns, by comparing names.

This option allows names that are commonly used for the Case column in a data source to be defined. The Data map window will use these to try and automatically map the Case Column.

Time Column Names

When the Data map window is opened, variables are automatically mapped to data columns, by comparing names.

This option allows names that are commonly used for the Time column in a data source to be defined. The Data map window will use these to try and automatically map the Time Column.

Weight Column Names

When the Data map window is opened, variables are automatically mapped to data columns, by comparing names.

This option allows names that are commonly used for the Weight column in a data source to be defined. The Data map window will use these to try and automatically map the Weight Column.

Histogram Bin Count

Defines the default number of bins (intervals) to be used when generating continuous histograms from the Data Explorer, Batch Query and Data Sampling windows.

Numeric format

Defines the default numeric format, used when displaying numeric data, in the Data Explorer, Batch Query and Data Sampling windows.

Maximum time

Defines the default maximum time (zero based) to be used in temporal (time series) queries. See also Maximum time type.

Minimum time

Defines the default minimum time (zero based) to be used in temporal (time series) queries. See also Minimum time type.

Terminal time

Defines the default terminal time (zero based) to be used in temporal (time series) queries. See also Terminal time type. The terminal time determines the point at which Terminal nodes link to temporal nodes.

Maximum time type

Defines whether the **Maximum time is an absolute time, or relative to the maximum evidence time. Times are zero based.

Minimum time type

Defines whether the Minimum time is an absolute time, or relative to the maximum evidence time. Times are zero based.

Terminal time type

Defines whether the Terminal time is an absolute time, or relative to the maximum evidence time. Times are zero based.

Sort discrete states by probability / lift / diff

Determines whether discrete queries in Query Explorer should be sorted by probability (or lift if the query has lift enabled).

Is Data Mode

When true the node, link and distribution buttons on the network tab of the main toolbar, launch the data driven equivalents Add nodes from data, Structural learning and Parameter learning respectively.

Parameter count warning level

Determines the point at which warnings are raised for a network with a large number of parameters.

Lift Gradient Cut Off

Determines the value at which comparison query arrows for lift have no transparency. The higher this value the higher the lift that is required for no transparency. For lift values less than one, the reciprocal of this value is used.

Sample count

Determines how many samples will be generated for each query, by an approximate inference algorithm that uses sampling. When not specified the algorithm will decide how many samples to use.

Data progress interval

The number of cases or temporal rows processed between progress updates, during data operations such as parameter learning or batch queries.

Show Start Page On Startup

When true, the start page is shown on application start up.