Filters
Processing guide
Build a reusable result table that contains exactly the records and columns a NIM workflow needs.
A filter starts with collected Vault data. You choose a start table, decide which records qualify, shape the result, and preview it before saving. Mappings, roles, name generators, jobs, and apps can then use that result.
Follow the configuration flowDirect link to Follow the configuration flow
| Stage | Decision | Where to go |
|---|---|---|
| 1. Start | Which collected table has one row for each person, account, group, or other object you want to evaluate? | Create and test a filter and select its Start Table. |
| 2. Choose records | Which rows qualify? Do you need a value comparison, a configured relationship, a check against another table, or to subtract another filter's population? | Use expressions, relations, lookups, or Exclude by Filter. Combine filter items with AND and OR when needed. |
| 3. Shape the result | Which columns and rows should the consumer receive? | Select and order columns, calculate values with JavaScript columns, group, sort or exclude rows, or append another filter. These are optional choices, not required steps. |
| 4. Check and save | Does the Result pane contain the expected records, columns, and values? | Preview normal records, expected non-matches, blanks, and duplicates; then save the filter. |
| 5. Use and maintain | Which mapping, role, job, or app will consume the result? | Connect the saved filter to that workflow. See manage and reuse filters when its requirements change. |
The table describes a useful configuration order. It does not claim that NIM executes every optional feature in that order. Start with a simple filter, preview each change, and add only the options the result needs.
Choose the right way to select recordsDirect link to Choose the right way to select records
| If you need to… | Use… | Effect on the result |
|---|---|---|
| Compare a column to a value, pattern, or supplied parameter | An expression item | Keeps rows that satisfy the condition; adds no columns. |
| Use related records or bring columns from another table into the result | A relation item | Joins connected data using a configured relation. A No match can find rows without a related record. |
| Check whether a value exists in another system table without joining its columns | A lookup | Keeps or removes rows based on the match. |
| Remove records related to another saved filter with a different start table | Exclude by Filter | Removes destination rows related to that filter's result. |
| Follow nested groups or a parent/child hierarchy | A recursive relation | Follows more than one level of a configured relation. |
An expression compares values already available to the filter. A relation needs configured table keys and a relation. A lookup compares a filter column with a field in another table and does not require a configured relation.
Example: find employees without a target accountDirect link to Example: find employees without a target account
- Create a filter with the collected employee table as its Start Table.
- Add an expression for the active employees the workflow should consider.
- If an employee-to-account relation exists, add a relation item with No matching target account. If you only need to compare an identifier to a target table field, use an exclude lookup instead.
- Select the columns the downstream mapping needs, including stable identifiers. Preview a known employee with an account, one without an account, and one with incomplete data.
- Save the filter after the Result pane contains only the intended employees. Then use it as the input to the provisioning mapping.
Use the relation or lookup approach that matches your collected data and keys. You do not need both for the same account check.
Reuse and troubleshoot filtersDirect link to Reuse and troubleshoot filters
Child filtersDirect link to Child filters
A child filter inherits its parent's criteria, which appear locked, and adds its own restrictions. The child can adjust copied columns, calculated columns, sorting, and order independently. Use one for a narrower population based on shared rules, such as Accounting leads derived from an Accounting employee filter. See create and maintain child filters.
Filter exclusionsDirect link to Filter exclusions
Use Exclude by Filter for mutually exclusive populations drawn from different, related start tables. It cannot exclude another filter with the same start table. For a narrower population that shares the same start table and criteria, consider a child filter. For row-level options such as removing duplicates or Boolean-flagged rows, see sort and exclude rows.
Invalid filtersDirect link to Invalid filters
An invalid filter is underlined in red and returns invalid output. A mapping, role, or other object that depends on it cannot produce the expected result until the filter is corrected.
Check the start table, expressions, relations, selected columns, and collection status first. See A filter is marked invalid. If every filter becomes invalid at once, check the license status before changing individual filters.
For a slow filter, read its performance statistics alongside the result preview and collection history.