1. Using Advanced Controls
  2. Model Filter Conditions

Model Filter Conditions

A model filter specifies an attribute of an object, then selects records with attribute values that satisfy a condition. You select the condition and elements that complete it: one or more constant values, or another attribute of an object.

For example, this is a filter that uses a condition involving a constant value: the Payment Amount attribute of the Payment business object is greater than 5,000 dollars.

Select among the following conditions as you create filters for transaction models. Each condition is available only to attributes it's appropriate for.

  • Mathematical operators: The filter returns results if the value of one attribute equals, doesn't equal, is less than, is less than or equal to, is greater than, or is greater than or equal to either a constant value or another attribute value. For a date attribute, "less" means "earlier," and "greater" means "later."

    You can set an attribute of a business object equal to itself. The filter returns groups of records. In each group, the attribute equals a specific value. If the attribute were Supplier ID, for example, all records for supplier ID 1234 would form one group, all records for ID 2345 would form another group, and so on.

  • Similar and Similar to: The Similar condition checks for similarity in the values of one attribute. The Similar to condition checks for similarity in two attributes; a value of either may be similar to other values of the same attribute or to values of the other attribute.

    For either condition, define a standard of similarity. The filter then collects records into groups. In each group, attribute values meet your standard in a distinct way.

    • Text: Enter a percentage. Strings are similar if the number of characters they share is at least that percentage of total characters. For example, six-character strings are 50 percent similar if they contain three matching characters.

      Values with distinct sets of matching characters form distinct groups. For example, one group might contain strings with abc, and another strings with xyz.

      Characters in one string need not be consecutive to match characters in another string. By default, the filter checks whether entire strings are similar. You can select a Similar Word advanced option to check for similarity in any single word in each string.

    • Number: Enter a percentage. This sets a lower limit. An upper limit is the same number of points above 100 percent as your percentage is below. For example, if you enter 85 percent, you establish two limiting values, 0.85 and 1.15.

      The filter takes the average of numbers already included in a group, multiplies it by the limiting values to create a range, and admits a new value if it falls within the range. Each time a new value is added, the group average changes, and so does the range for admitting new values.

      In the 85 percent example, one group might include 22 and 20. Another might include 9 and 8. Although 7 is one apart from 8, just as 8 is from 9, 7 wouldn't be similar to 9 and 8. That's because the average of 9 and 8 is 8.5, the 85 percent range around 8.5 is 7.225 to 9.775, and 7 falls outside that range.

    • Date: Enter a number of days. Dates are similar if they fall within this span.

    By default, a record belongs only to the group it qualifies for first. (Text strings are evaluated in ascending alphabetic order; numeric values in descending numeric order.) You can set a Generate Results for Similar Groups advanced option to have records belong to all groups they qualify for.

    The filter excludes records that don't qualify for any of the groups it creates. You can select an Include Unique Data Rows advanced option to get records the filter would otherwise exclude.

  • Different than: Get records that the Similar to condition would exclude. Criteria for these two conditions are the same.

  • Between: The filter returns results if the value of an attribute falls between two constant values that you select.

  • Is blank and Is not blank: The filter returns records that either have no value, or any value, for a specified attribute. The filter consists only of the attribute and the Is blank or Is not blank statement, because these two elements are sufficient to define the filter.

  • Is not related to: Specify two business objects to identify records existing in one, but not the other. For example, an invoice that's never been on hold should not appear in the Payables Invoice Hold object. So the filter "Payables Invoice Is not related to Payables Invoice Hold" returns records of invoices that have never been on hold.

  • Contains and Does not contain: A Contains filter returns results if the value of a text attribute includes either a text string you specify or a value of another text attribute you identify. A Does not contain filter returns results if the value of a text attribute doesn't include specified text; in effect, it returns all records that a Contains filter wouldn't.

  • Matches any of and Matches none of: The filter returns results if the value of an attribute is a text string, number, or ID that matches any in a specified set of values, or matches none of them. Use either a semicolon or a comma to delimit values. The match must be exact. (The Matches any of condition behaves a little differently in a filter that uses the Inclusive or Exclusive function. See Create a Function Filter.)

  • Starts with and Ends with: The filter returns results if the value of an attribute begins or concludes with a specified string of alphabetic or numeric characters. You may specify a constant value or another attribute that returns text values.

  • Expresses: The filter finds records in which the value of a specified attribute is a text string containing any term in a dictionary. This dictionary must exist as an imported object. Supply the name of that imported object in the Object field to the right of the Expresses condition, and Word in the Attribute field. (See Create an Object as a Dictionary for the Expresses Condition.)

  • Related to: Establish a join relationship between an attribute of a user-defined object and an attribute of any other delivered, imported, or user-defined object. Or, establish a join relationship between an attribute of a "stand-alone" object and an attribute of any other delivered object. (A stand-alone object is a delivered object with no default relationship to other objects.) Specify the attribute of the user-defined or stand-alone object to the left of the Related-to condition, and the attribute of the other object to the right.

    This join relationship is valid only for the model or control it exists in. A user-defined or stand-alone object can have only one dynamic join relationship per model. If the attribute to the left of the condition is a date, the attribute to the right must also be a date; otherwise, attribute types need not match. Use the Business Object Visualizer to determine whether a business object is stand-alone.