Skip to end of metadata
Go to start of metadata

The Target Attributes pane allows you to change attributes for the columns and tables in the SuperSTAR database.

To display the Target Attributes pane, click the icon on the toolbar:

The attributes shown will depend on what is currently selected in the Target View. For example, different attributes are shown if the currently selected item is a database, fact table, or column.

Database Attributes

These are shown when you select a database in the Target View:

The database attributes for a target schema are available at any time during the design of a target schema.

AttributeDescription
Label

The database label. By default, this is set to the target database filename (without the .sxv4 extension).

You can change the label if you wish, although this label is not used in the SuperSTAR clients (you can set the display name for the database to use in the clients when you add the database to the catalogue in SuperADMIN).

NameThe database name. This is set automatically based on the target database filename and cannot be changed.
LocationThe location of the target database file. This cannot be changed.

Whether to generate Rkeys automatically for this database. See Automatically Generate R Keys for Perturbation - SuperCHANNEL for more details.

The Partial Build setting determines the number of rows to include when building the database:

  • If the Partial Build check box is selected then only the specified number of rows from the source database will be used for the build.
  • If the Partial Build check box is not selected then all rows in the source database will be used for the build.

This option is designed to allow you to test your configuration without having to build the entire SXV4.

It is particularly useful if you have a very large source database. You can configure SuperCHANNEL to run a partial build with just a small number of rows to check that you are happy with the output. Once you are satisfied with the design of the SXV4 you can turn off the partial build and build the entire database.

When using the Partial Build option you may see errors in the logs indicating that some of the foreign key constraints did not succeed. This is typically because one of the child tables references a row in a parent table that was not included in the partial build.

The number of records channelled before changes are committed to the target database.

This setting affects disk usage. See Resource Considerations - SuperCHANNEL for more information.

There is also a Commit Points setting for each fact table. If this is configured at the fact table level in addition to the database level, then the fact table setting takes precedence over the setting at the database level.

The maximum number of cleansing actions allowed for a database build.

If you select the Stop on Error check box, then the build process will stop if it reaches the specified number of cleansing actions.

The default summation used during tabulation. This can be either:

  • A measure column from a fact table (e.g., income, account profit, average account balance).
  • A row count of a fact table (e.g., the number of customers of accounts).

The drop-down list shows all the fact tables and measure columns available in the database.

You can also tell what is currently the default summation by looking for the + icon in the Target View:

The capitalisation of the table and column labels:

NoneDo not change the case of the table and column labels.
LowercaseChange all labels to lower case.
UppercaseChange all labels to upper case.
HeadingChange all labels so that each word starts with a capital letter.

When you select one of the options, all the target labels will be changed immediately. If you subsequently edit the labels, then your chosen capitalisation option will not be applied to your modified labels automatically.

If you change some of the labels and want to reapply your chosen label formatting, you must go back to the Target Attributes and reselect one of the formatting options.

The default grouping mode for the target database.

Flat
  • All fact table measure columns will be added to the top level of the Summations folder.
  • Any multi-response fact tables will be added to a Multi-responses sub group.
Hierarchy
  • All fact table counts will be added to the top level of the Summations folder.
  • All fact table measure columns will be added to a sub group named after that fact table.
  • All multi-response fact tables will be added to a sub group named <TABLE> Multi-responses (where <TABLE> is the name of the fact table the multi-response table references). This multi-response group is then placed under a group with the name <TABLE>.

  • All multi-response fact table columns are added under a group named after the fact table the multi-response table references.

The selected grouping mode will be applied by default. You can manually modify the grouping by editing the settings in the Grouping pane. If you use the Reset Grouping option in the Grouping pane then the groupings will be reset to either flat or hierarchical based on this setting.

Table Attributes

These are shown when you select a table in the Target View:

AttributeApplies ToDescription
LabelAll table types

The display name for this table.

  • For fact tables, this will be shown in the SuperSTAR clients.
  • For classification tables, the label will not generally be shown in SuperSTAR clients, although it is a good idea to label a classification table so that it is easy to identify the fact table column that refers to it.

    Classification table labels are used in SuperWEB2, in the Select all at level drop-down list (although you can also change that text by configuring SuperWEB2).

  • Plain tables are not included in the target database, so any label you set for these tables will not appear in SuperSTAR clients. Setting a descriptive label for these tables may help readability of the target schema.
NameAll table typesThe table name. This is based on the source database and cannot be changed.
SourceAll table typesThe name of the source table in the source database. This cannot be changed.
UsageAll table types

The table type (Fact, Classification or Plain).

ConditionAll table types

A SQL WHERE or ORDER BY clause that will be executed when the database build is run. For example, you might use this to sort your classification tables.

  • Only drivers that support selection queries can use this function.
  • Each JDBC or ODBC driver supports its own specific SQL formatting such as quote characters. You must ensure that clauses specified as conditions are in the correct format.

The following example shows a compound WHERE clause:

Any references to columns must use the column Name attribute, not the Label:

Primary Key SequenceFact tables only

The sequence of primary keys, if the fact table has more than one primary key.

Select one of the keys in the list and click the up or down arrow to move it up or down.

The primary key sequence is important for performance reasons:

  • To increase channelling performance when a primary key extends across multiple columns, sort the source data in primary key order before loading.
  • When the source database table is ordered according to the compound primary key and the correct order of the primary key columns is specified for the target SXV4 table there may be increased channelling performance.
  • Specifying the order of primary key columns in a compound primary key enhances channelling performance when inserting records into tables.
  • SuperCHANNEL channelling performance may decrease if tables with compound primary keys have to be checked for record integrity. It is best to avoid using compound primary keys where possible.
Commit PointsFact tables only

The number of records channelled before changes are committed to the target database.

If you set this value at the fact table level, it will override any commit points configured at the database level.

Multi-ResponseFact tables only

Whether or not the fact table is a multi-response table.

Select the check box to configure the fact table to be a multi-response table.

Multi-response tables are indicated in the Target View with a multi-table icon:

Column UsageClassification tables only

The names of the columns used for the Code and Name.

All valid classification tables must have one column whose usage is set to Code, and another column whose usage is set to Name.

You cannot edit this from the table Target Attributes pane. To change the column usage you need to select the individual columns and specify which one is which.

Totals AppropriateClassification tables only

Whether to allow totals for this classification table. This setting is designed for situations where it does not make sense to generate totals based on this classification (for example, you may have a time series database where it does not make sense to sum across years). The setting currently applies to SuperWEB2 only; it does not affect totals in SuperCROSS:

  • Select the check box to allow totals (this is the default).
  • Clear the check box to disallow totals. Users of SuperWEB2 will not be able to create a total for a field that is based on this classification table.
Rkey GenerationFact tables only

The seed and generator to use when generating Rkeys for this fact table. These settings only appear if the Generate Rkey Automatically check box is selected. See Automatically Generate R Keys for Perturbation - SuperCHANNEL for more details.

Plain tables are not included in a database build. The attributes for a plain table are not useful until the table has been redefined as either a fact or classification table.

Column Attributes

These are shown when you select a column in the Target View:

AttributeDescription
Label

The display name for this column.

  • For fact table columns, this will be shown in the field list in the SuperSTAR clients.
  • For columns in classification and plain tables, the label will not be shown in SuperSTAR clients, but setting a descriptive label may help readability of the target schema.
NameThe column name. This is extracted automatically from the source database and cannot be changed.
IndexThe column index from the source database. This cannot be changed.
SizeThe column size from the source database. This cannot be changed.
Usage

The column usage.

Measure

Measures are values that can be used in calculations. They are generally continuous variables such as income (which are not suitable as classifications).

You can only set a column as a measure if its data type is numeric (e.g., integer, double, or bigint)

Classified

A column in a fact table that has a reference to a classification table. This is only applicable to columns with non-numeric data types.

When a fact table column is set to Classified, the Cleansing actions are available (see below).

Plain

A plain table column.

To reduce build time and the size of the target database, any columns in a classification table that are not either the Code or Name columns or a foreign key to another classification table, should be set to Plain and excluded from the build.

Code

A column in a classification table that contains the set of codes for this classification.

  • It must be the primary key for the classification table, and linked to the fact table column using a reference.
  • Each classification table can have a maximum of one Code column.
  • A column cannot be set to Code if its data type is DOUBLE, DATE, TIME, TIMESTAMP, or OTHER.
NameThe classification table that contains the names of the values for this classification. These are the values displayed in the SuperSTAR clients.
Data Type

The column data type. You can change the data type by selecting a new value from the drop-down list:

Primary Key

Indicates whether or not this column is a primary key. You can also identify which columns are primary keys by looking for the key icon in the Target View.

Primary keys are used when linking tables. For example, if there are multiple fact tables, one of the fact tables must has a foreign key that is linked to the primary key of the other fact table.

In addition, each Classified column in the fact tables must be linked to the primary key of its classification table (the Code column).

Included in Build

Indicates whether or not the fact table column is included in the build.

Excluded columns appear in the Target View but are greyed out:

Any column with the data type OTHER is automatically excluded from the build. You will need to change the column data type in order to include it in the build.

You should always exclude columns from the build if their data is not required in the target database. This helps to optimise build times and keep database sizes to a minimum.

Cleansing

The action to take when the build fails to match a value in the fact table column with the set of possible values in the linked classification table.

For example, the Gender column might be linked to a classification table containing the codes M (Male), F (Female), and U (Unknown). What should SuperCHANNEL do if it encounters an empty value for this column in one of the records in the fact table?

You can choose from the following cleansing actions:

Cleansing ActionDescription
Add To Classification

Add the new value to the classification table. This is the default action for classified columns.

This option is only available for classified columns; it cannot be used for foreign key columns to other fact tables or between a fact table column and a multi-response fact table. If you try to set these to use the add to classification action then SuperCHANNEL will automatically revert to skip.

BinConvert the value to a specified "bin" value. The bin value must already exist in the corresponding classification table.
NoneDo not capture exceptions. Use this action for non-SXV4 drivers.
Skip Row

Skip this record.

This can be a useful cleansing action for a column linking two fact tables. For example if you want SuperCHANNEL to skip all Account records that do not have associated Customer records.

Use caution when selecting this option: your target may end up with fewer fact records than the source.

Stop

Stop the build process.

This is the default action for foreign key columns to other fact tables.

 

Plain table columns are not included in a database build. The attributes for a plain table and its columns are not useful until the table has been redefined as either a fact or classification table.

  • No labels