Cube Management: Map Data and Define Relationships

Overview

The Mapping tab adds structure, consolidations, and cleanup on top of your raw transactional data. It is where you:

• Group detailed items into families, brands, or categories.
• Normalize messy naming (Walmart vs Wal-Mart becomes Walmart).
• Attach extra attributes you do not want to repeat on every row of the fact data.
• Build compound keys when a single column is not enough to uniquely identify a record.

Think “star schema.” Your Source is the fact table. Mapping tables are lookups you join to the fact on a key.

What you see on the Mapping tab

• A list of Cube Mapping Tables (example: Item List, Customer List).
• Icons on each row:
  • ⟳ Replace table (looks like a refresh symbol, but it replaces the mapping file
  • ⤓ Download to get the current mapping file to Excel
  • ⋮ More menu copens actions: Edit Table Link, Remove Table Link, Delete Table, Upload Center.

Upload Center: This shows all previous mapping files you’ve uploaded, giving you a lightweight form of version control. You can review past versions and re-upload an updated file as the new active version.

• Add Mapping Table button opens Add Mapping Table with three options:
  1. Upload an existing file
  2. Build a new file from source data (export unique keys)
  3. Select an existing table (reuse a mapping already in the project)

Key concepts

• Source Table Column: the key in your fact data (for example, ItemID or a combined Company - Customer).
• Mapping Column: the matching key in the mapping table (for example, ID in Item List).
• One row per key in the mapping table. Duplicates create ambiguous joins.
• Many fact rows to one mapping row. Do not try to map one fact row to multiple mapping rows.
• Attributes live in the mapping table. Your fact table stays lean.

Add or link a mapping table

Option A: Upload an existing file

Use this when you already maintain a clean lookup in Excel or CSV.

1. Click Add Mapping Table → Upload an existing file → Continue.
2. Upload the file. Ensure it has a key column that matches a fact key.
3. After upload, choose Edit Table Link.
4. In Edit Mapping Table, pick the Source Table (your fact) and select the Source Table Column that serves as the key.
5. Pick the Mapping Table and select its Mapping Column that matches the key.
6. Confirm.
7. Click Process cube data.

Good file shape

• One header row.
• A single unique key column.
• Additional columns are attributes you want available in pivots (Category, Brand, Parent Customer, Region, Launch Date, etc.).

Replace vs. Append

• Replace is the only supported method in the UI.

• If you need to add new keys, edit the file in Excel (append offline), then upload as a replacement.

Option B: Build a new file from source data

Use this to generate a clean starting template directly from your cube.

1. Click Add Mapping Table → Build a new file from source data → Continue.
2. Choose the Source Table Column you want to map (for example, ItemID or Company - Customer).
3. Download the generated file. It contains one row per unique key from the fact data.
4. Open in Excel, fill in the attributes you want (Category, Channel, etc.).
5. Back on the Mapping tab, use Upload Center and Replace Table to upload your edited file.
6. Open Edit Table Link, confirm the key-to-key link, then Process cube data.

Why this is useful
It guarantees your mapping file has every key that appears in the facts, which reduces “unmapped” records.

Option C: Select an existing table

Use this to reuse a mapping table already maintained in the same project.

1. Click Add Mapping Table → Select an existing table → Continue.
2. Choose the mapping table to link (for example, reuse a master Customer List across multiple cubes).
3. Open Edit Table Link, pick the source key and the mapping key, then Confirm and Process cube data.

Benefit
One golden lookup shared by many cubes means less maintenance and consistent reporting.

Create combined keys (compound joins)

Sometimes a single column is not unique. For example, a customer name might repeat across companies. Use Create Combined Columns to build a compound key like Company - Customer.

Steps

1. Open Edit Mapping Table.
2. Click Create Combined Columns under Source Table Column.
3. In Add Column, name it clearly (example: Company - Customer).
4. Pick the two or more string columns to concatenate (example: CompanyName and CustomerName).
5. Confirm.
6. Repeat on the Mapping Table side if needed, building the same combined key.
7. Select the new combined column on both sides as your link, then Confirm and Process cube data.

Tips

• Keep combined-key columns set to string in Configuration.
• Use a consistent separator in both files, for example Company - Customer.
• Avoid leading and trailing spaces in any component columns.

⚠️ Note on the Advanced toggle: Most users will only have access to Basic combined columns. If you need more complex combined keys (e.g., calculated expressions or logic-based concatenations), please reach out to Support or your project admin for assistance.

Edit, replace, or remove a mapping

• Use the ⋮ More menu → Edit Table Link to change the key pairing.
• Use Replace link to swap out for a different mapping file.
• Use Remove Table Link to unlink without deleting the file.
• Use Delete Table to remove the file completely from the project.
• Download anytime to audit or modify the mapping offline.

Common patterns and examples

• Items
  • Key: ItemID → Item List.ID
  • Attributes: Brand, Family, Size, UOM, Color, Launch Date.
• Customers
  • Key: Company - Customer (combined) → Customer List.Company - Customer
  • Attributes: Parent Customer, Channel, Region, Sales Rep, Status (Active/Inactive).
• Calendar enrichments
  • Key: CalendarDate → Date Table.Date
  • Attributes: Fiscal Month, Fiscal Year, QTR, Week Number, YTD flag.
• Channel or Market
  • Key: ChannelCode → Channel Map.Code
  • Attributes: Channel Group, End Market, eCom vs Retail.

Quality checks

• Zero duplicates in the mapping key column. One row per key only.
• Zero blanks in the mapping key column.
• No accidental spaces. Trim leading and trailing spaces in keys and attributes.
• Consistent case. Prefer all caps or proper case.
• Data type alignment. Keys used to build combined columns must be strings.
• Coverage. If many fact rows are “unmapped” in Excel, use Option B to regenerate the key list, compare, and fill gaps.

Processing and seeing results

1. After linking or updating a mapping, click Process cube data.
2. When processing completes, open Excel and refresh your cube.
3. New attributes from the mapping appear as fields for slicing and filtering.
4. If processing fails, return to Mapping and check for duplicate keys, wrong data types, or a broken link definition.

Best practices

• Keep mapping files small and tidy. Only include attributes you plan to use.
• Use Replace for controlled updates, not Append, to avoid duplicate keys.
• Prefer shared master lists (Option C) for customers, items, and dates across cubes.
• Name keys clearly. Company - Customer reads better than CompCust.
• Version your files in your team’s shared drive. Add a short change log tab.
• Schedule a monthly review to retire stale values and add new ones.

Troubleshooting quick list

• Keys not matching → verify the exact same text and separator on both sides.
• Missing combined column → create it on both Source and Mapping, confirm string type.
• Lots of “unmapped” records → use Build a new file from source data, compare, and fill.
• Processing green but fields missing in Excel → refresh the workbook or re-download a fresh cube.

Related and next up

• Before this: Cube Management: Loading and Managing Data
• Next article: Cube Management: Configuring Key Fields and Data Types
• Also helpful: Cube Management: Calculations and Adding Measures

Mapping is where your cube starts to feel customized for your business. By linking clean Cube Mapping Tables to your raw fact data, you can group products, clean up customer names, or enrich transactions with extra attributes that drive better analysis. Remember that mapping is iterative — it will likely evolve as your data and reporting needs do.

Once your mappings are in place and processed, you’re ready to move into Configuration, where you’ll set data types, key fields, and measures that determine how your cube calculates and displays results.