You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 6 Next »

Model Documentation

The WhereClause class is used to define data selection criteria that are used for the selection and grouping of data for analysis.

Error rendering macro 'excerpt-include'

No link could be created for 'ARSP:Class Diagram: WhereClause'.

There are 3 specializations of the WhereClause class that allow an identifier value and label to be assigned for the defined selection criterion:

  • The AnalysisSet class is used to define subject populations (i.e., criteria for the selection of subjects for inclusion in an analysis), as described in the AnalysisSet section.
  • The DataSubset class is used to define criteria for the selection of data records for inclusion in an analysis, as described in the DataSubset section.
  • The Group class, through use of its two specializations, AnalysisGroup and DataGroup, is used to define the criteria for grouping subjects and data records (respectively) for analysis, as described in the DataGroupingFactor section.

Each instance of the WhereClause class (or any of its specializations) defines either:

  • A simple condition based on the value or values for single dataset variable or
  • A compound expression that either:
    • combines selection criteria, which are defined in other instances of the WhereClause class (or its specializations) or
    • negates the selection criteria defined in another single instance of the WhereClause class (or its specializations) .

Compound expressions can combine or negate simple conditions and other compound expressions.

In each instance of the WhereClause class,

  • The level attribute contains an integer that indicates the level at which the specified condition is being applied, with a value of "1" indicating the "top" level.
    • The level attribute has a value of "1" in any instance of a specialization of the WhereClause class where an identifier value is assigned. 
    • The level attribute contains a value greater than 1 only in instances of the WhereClause class that are used to define sub-conditions within a compound expression.
  • The order attribute contains an integer to order this instance of the WhereClause class with respect to other other instance of the WhereClause at the same level.
    • When the value of the level attribute is greater than 1 (i.e., this instance of the WhereClause class defines one of the selection criteria being combined in a compound expression), the order attribute defines the order of this instance of the WhereClause class with respect to the other selection criteria with which it is being combined.
    • When the value of the level attribute is "1":
      • In instances of either of the specializations of the Group class, the order attribute is used to define the order of the group with respect to other groups in the grouping factor.
      • In all other situations, the order attribute should have a value of "1".

In an instance of the WhereClause class that defines a simple condition, the condition attribute contains an instance of the WhereClauseCondition class in which:

  • The dataset attribute contains the name of the dataset in which the condition is being applied.
  • The variable attribute contains the name of the variable (within the specified dataset) to which the condition applies.
  • The comparator attribute contains a value from the ConditionComparatorEnum enumeration ("EQ" for "equal to", "LT" for less than, etc.) to indicate how values of the specified dataset variable are compared with the specified selection value(s).
  • The value attribute specifies the selection value, or values, for comparison with values in the specified dataset variable. If the value of the comparator attribute is:
    • "IN" or "NOTIN", there should be at least 2 values specified in the value attribute; otherwise there should be no more than 1 value specified in the value attribute.
    • "EQ" or "NE", the value attribute may contain no value to indicate a condition based on a missing value (i.e., "is missing" or "is not missing" respectively). 

Example

This example shows an instance of the WhereClause class (or one of its specializations) that defines a simple condition based on a single specified selection value. The example shown specifies that the value of the SAFFL variable in the ADSL dataset is equal to "Y", which could be expressed as:
  • ADSL.SAFFL EQ 'Y'

Note that some attributes (such as the expected level and order attributes) have been excluded from this example. 

YAML Example
...
condition:
  dataset: ADSL
  variable: SAFFL
  comparator: EQ
  value:
  - Y

This condition could be represented in tabular form as:

$titleHtml
datasetvariablecomparatorvalue
ADSLSAFFLEQY
$warningHtml
Other conditions based on a single value (i.e., with a comparator value of "EQ", "NE", "LT", "LE", "GT", or "GE") would be represented in a similar way.

Example

This example shows an instance of the WhereClause class (or one of its specializations) that defines a simple condition based on multiple specified selection values. The example shown specifies that the value of the AEREL variable in the ADAE dataset matches 1 of the listed values ("POSSIBLE" or "PROBABLE"), which could be expressed as:
  • ADAE.AEREL IN ('POSSIBLE','PROBABLE')

Note that some attributes (such as the expected level and order attributes) have been excluded from this example. 

YAML Example
...
condition:
  dataset: ADAE
  variable: AEREL
  comparator: IN
  value:
  - POSSIBLE
  - PROBABLE

This condition could be represented in tabular form as shown in the following table where the multiple selection values are represented as a pipe (|)-delimited list:

$titleHtml
datasetvariablecomparatorvalue
ADAEAERELINPOSSIBLE|PROBABLE
$warningHtml
Other conditions based on multiple selection values (i.e., with a comparator value of "IN" or "NOTIN") would be represented similarly.

Example

This example shows an instance of the WhereClause class (or one of its specializations) that defines a simple condition based on a missing value. The example shown specifies that the value of the BASE variable in the ADVS dataset is not missing, which could be expressed as any of the following:
  • ADVS.BASE NE ''
  • ADVS.BASE IS NOT MISSING
  • ADVS.BASE IS NOT NULL

Note that some attributes (such as the expected level and order attributes) have been excluded from this example. 

YAML Example
...
condition:
  dataset: ADVS
  variable: BASE
  comparator: NE
  value:

This condition could be represented in tabular form as:

$titleHtml
datasetvariablecomparatorvalue
ADVSBASENE
$warningHtml
Other conditions based on a missing value (i.e., with a comparator value of "EQ" or "NE") would be represented similarly.







SelectionGrouping
SubjectsanalysisSetsanalysisGroupings
DatadataSubsetsdataGroupings

  • No labels