A dimension column tells Cube which dimension member each source value belongs to, so a location called "Remote Worker - N/A" at the source rolls up to the member you report on in Cube. Mapping lives on the table, one tab per dimension column, and doubles as the filter for dimension values: anything you map to Ignore stays out of the table.
What you can use it for
- Rolling source data up to your dimensions. Map HRIS locations, departments, or cost centers to the members you report on.
- Keeping records out. Map a value such as Contractors or Interns to Ignore and those records are never imported.
- Reusing a mapping. Download the mapping from one table and upload it to another table on the same integration.
- Catching new values. When the source adds a location, the next Import stops and the table tells you how many values need mapping before anything is written.
Before you start
- Editing a mapping needs Integrations access. Without it you can open the Mapping page read-only.
- Mapping starts blank on every new table. Nothing is copied from the integration's own mapping or from any other table.
- Tables have no mapping rules and no Mapping Assistant. You map values by hand in Edit Mapping or by uploading a CSV.
- Saving a mapping writes nothing to the table. The new mapping applies the next time you run Import.
How to map values
1. Open the connected table and select the Mapping icon in the header, next to the filter count. The Mapping page opens with one tab per dimension column and the count of unmapped values on each.
2. Select Edit Mapping.
3. On Map Dimensions, select the Cube dimension this column maps to under Cube Dimension, then select Confirm. You do this once per tab.
4. For each source value, select a dimension member under Cube Dimension, or select Ignore.
5. Switch to the next tab and repeat.
6. Select Save Mapping. Cube confirms how many dimensions were mapped.
7. Return to the table and select Import to apply the mapping.
Ignore counts as mapped. Once every value on every tab is mapped or ignored, the table is ready to import.
How to reuse a mapping on another table
1. On the Mapping page, open Settings (the gear icon) and select Download Mapping. Cube downloads the current tab as a CSV with two columns: the source value and the Cube member.
2. Open the other table's Mapping page, select the same dimension tab, and select the Upload icon.
3. Choose the CSV. Cube maps every value it finds and tells you how many were mapped.
Use Ignore in the Cube member column to ignore a value. The two tables must be on the same integration so the source values match.
How Cube keeps the mapping current
Cube reads the source's dimension values when the table is created, when you run a dimension sync from the Connection Overview, and at the start of every Import. New values appear on the Mapping page as unmapped, and the filter on the page lets you show only those.
If an Import finds unmapped values, it stops before writing anything and the Import log records a Failed run with that reason. Map or Ignore the new values and run Import again.
What happens to rows when a mapping changes
| Change | On the next Import |
|---|---|
| Unmapped value mapped to a member | Records with that value come in as Synced rows carrying the member |
| Member changed to another member | Synced rows with that value carry the new member |
| Member changed to Ignore | Synced rows with that value move to Out of Sync. They keep their previous value in that column and all their custom columns, and no longer roll up to a member. |
| Ignore changed to a member | Records that were kept out come in as Synced rows. Out of Sync rows with that value return to Synced with the new member, custom columns intact. |
| A value Ignored before the first Import | Those records are never brought in, and nothing shows as Out of Sync |
Renaming a dimension member in Cube needs no Import at all: the table shows the new name immediately, because the mapping points at the member, not its name.
Other things to know
- Mapping is per table. Saving on one table changes no other table, even on the same integration.
- Switching a column from a dimension type to Text in Table Settings removes that dimension's tab and its mapping after you save.
- Filter rules and Ignore do the same job for different column types. Filter Text, Numeric, and Date fields with rules; filter dimension fields with Ignore.