Overview
The Cases with Unchanged Attribute filter selects cases where a specified attribute maintains the same value throughout the entire case. This case-level filter examines all events within each case and returns only those cases where every event has an identical value in the selected attribute. This filter is particularly useful for identifying processes where certain properties remain constant, detecting data quality issues, or finding cases that follow standardized patterns without variation.
Common Uses
- Data Quality Validation: Identify cases where attributes that should change (like status or location) remain static throughout the process, potentially indicating data recording issues.
- Process Standardization: Find cases that were handled by a single department, resource, or system without handoffs or transfers.
- Consistency Analysis: Detect cases where key attributes like priority level, customer type, or product category remained unchanged from start to finish.
- Single-Location Processing: Identify cases processed entirely at one location or facility without transfers between sites.
- Dedicated Resource Analysis: Find cases handled by a single resource or team member throughout the entire process.
- Static Configuration Detection: Discover cases where configuration values or system settings remained constant, indicating stable processing conditions.
Settings
Event Attribute Name: Select the event attribute you want to check for consistency. The filter will return all cases where every event has the same value in this attribute. The attribute must exist in your event table and be of a supported data type (String, Integer, DateTime, TimeSpan, Decimal, Boolean).
Note: The filter uses exact value matching. For numeric and date attributes, values must be identical. For string attributes, the comparison is case-sensitive. Null values are treated as valid - if all events have null values in the selected attribute, the case will be included in the results.
Examples
Example 1: Finding Single-Department Cases
Scenario: You want to identify purchase orders that were processed entirely within a single department without any inter-departmental handoffs.
Settings:
- Event Attribute Name: "Department"
Result: The filter returns only cases where every event has the same department value, such as all events having "Finance" or all events having "Procurement".
Insights: These cases represent streamlined processes without departmental transfers. This can help identify:
- Departments with end-to-end process ownership
- Cases that avoided coordination overhead
- Potential best practices for process efficiency
- Simpler cases that didn't require cross-functional collaboration
Example 2: Detecting Unchanging Priority Levels
Scenario: You need to find cases where the priority level never changed from start to finish, which could indicate either straightforward processing or a failure to escalate urgent issues.
Settings:
- Event Attribute Name: "Priority"
Result: The filter selects cases where all events have the same priority value (e.g., all "Low", all "Medium", or all "High").
Insights: This reveals:
- Cases that maintained their initial priority classification
- Potential issues where urgent cases weren't escalated
- Standard processing patterns for different priority levels
- Opportunities to implement dynamic priority adjustment
Example 3: Identifying Static Status Cases
Scenario: Find cases where a status attribute never changed, which could indicate data quality problems or incomplete process execution.
Settings:
- Event Attribute Name: "Status"
Result: The filter returns cases where the status field has the same value in all events.
Insights: These cases may represent:
- Data recording errors where status updates weren't logged
- Cancelled or abandoned cases that never progressed
- System integration issues preventing status updates
- Cases requiring data quality remediation
Example 4: Single-Location Processing
Scenario: Identify manufacturing orders that were completed entirely at one production facility without transfers to other locations.
Settings:
- Event Attribute Name: "Location"
Result: The filter selects cases where all events occurred at the same location.
Insights: This can reveal:
- Products manufactured at a single facility from start to finish
- Cases that avoided logistics complexity
- Location-specific capabilities and expertise
- Opportunities for centralized processing models
Example 5: Consistent Resource Allocation
Scenario: Find cases where the same resource handled all activities, indicating dedicated case ownership.
Settings:
- Event Attribute Name: "Resource"
Result: The filter returns cases where every event was performed by the same resource or person.
Insights: These cases show:
- End-to-end case ownership by individual resources
- Potential efficiency gains from avoiding handoffs
- Resource specialization patterns
- Training opportunities for dedicated case handling
Example 6: Uniform Customer Type Processing
Scenario: Identify cases where the customer type classification remained constant throughout processing.
Settings:
- Event Attribute Name: "CustomerType"
Result: The filter selects cases where all events have the same customer type value (e.g., all "Premium", all "Standard", or all "Enterprise").
Insights: This helps understand:
- Customer segments with stable classifications
- Processes tailored to specific customer types
- Consistency in customer categorization
- Patterns in how different customer types are handled
Output
The filter returns a new dataset containing only the cases where all events have identical values in the specified attribute. Each returned case preserves all its original events and attributes. If a case contains any variation in the selected attribute's value across its events, that case is excluded from the results.
If no cases match the criteria (meaning all cases have at least one event with a different value), the filter returns an empty result set.
Technical Notes
- Filter Type: Case-level filter (removes entire cases, not individual events)
- Comparison Logic: Uses the first event's value as a reference and compares all subsequent events to it
- Null Handling: Treats null values as valid and consistent - cases where all events have null values are included
- Supported Data Types: String, Int32, Int64, DateTime, TimeSpan, Single, Double, Boolean
- Performance: Efficiently implemented using LINQ with early termination when a mismatch is found
This documentation is part of the mindzieStudio process mining platform.