Account
Skip to main content

Dimensional Accelerator Walkthrough

This walkthrough takes you from source metadata to a published star schema with the Dimensional Accelerator. It follows the walkthrough strip that the accelerator shows over its canvas, one step at a time, so you can read along inside the app. For every field, verb and dialog on the page, see the Dimensional Accelerator user guide.

Prerequisites​

Before you start, you need:

  • A target data mart project: a Project whose Target Connection is on the Data Mart integration stage. The accelerator derives the star into that target connection.
  • Source metadata to model: staging or persistent staging objects, or Data Vault hubs and links with their point-in-time tables. To import metadata from your own sources, use Import Metadata as the evaluation guide describes.

To start from a sample instead, follow the dimensional starting point of the evaluation guide: load a sample from the Sample Gallery, then choose the Dimensional flow card on Sample Home, which opens the Dimensional Accelerator. The 02 - MSSQL After Import sample holds the AdventureWorksLT tables as imported source objects with their columns, so the first thing you do is design. The 00.01 - MSSQL SSIS Starting Point sample holds the connections, batches and projects for a SQL Server warehouse, and you import the AdventureWorksLT source metadata yourself.

You can also open the accelerator from the Accelerator page with Open Dimensional, or from the Dimensional link in the model step of the first build checklist.

Step 1: Pick the Target Project​

The strip reads Step 1 of 3: Pick the target project and asks you to pick the data mart target project with the Project picker in the toolbar.

  1. When exactly one project qualifies, the accelerator selects it for you and the strip moves on to step 2.
  2. Otherwise, open the Project: picker in the toolbar and choose the target project.
  3. If the strip says it found no eligible projects, no project has a target connection on the Data Mart integration stage. Give the project whose objects you want to accelerate such a target connection, then reload the page.

The strip stays on this step until a target project is selected.

Step 2: Select Source Objects​

The strip reads Step 2 of 3: Select source objects and asks you to tick source objects in the tree on the left.

  1. Choose Add sources on the command bar. The source tree opens with the cursor in its search box.
  2. Tick the source objects for the star. The tree offers objects on the Staging Area, Persistent Staging Area, Data Vault and Business Data Vault stages. For a Data Vault, tick the hubs and links, not their satellites or point-in-time tables.
  3. Watch the canvas. As soon as you tick an object, the accelerator derives the star and draws a card for each dimension and fact, and the strip moves on to step 3.

Every source needs a role, Dimension or Fact. A source without a role lands on the canvas as a gray card marked Needs a role. Right click the card and pick Make this a dimension or Make this a fact. The card becomes that object in place. Once a card has a role, the Role field of its panel switches it between Dimension and Fact.

If your business model types business entities for the dimensional model, Start from the business model under the strip lists them. Tick the entities you want and choose Add to add their linked objects to the selection.

Step 3: Review and Publish​

The strip reads Step 3 of 3: Review and publish and asks you to review the star and the diagnostics, then Publish. While the derivation runs, it tells you that the star appears once derivation completes.

  1. Read the cards. A dimension card names its source under its name and carries a Type 1 or Type 2 chip for its history. A fact card names its grain from the dimensions it references and carries a chip that counts its measures. Click a card to point the panel on the right at it, and choose Object Details from the card menu to open that panel.
  2. Give every gray card a role. While gray cards remain, the strip reminds you that Publish writes only dimensions and facts.
  3. Shape the star where you need to. Rename a card or a column with a double click, set history with Track History and Track Changes, switch fact columns with Set as Measure and Set as Attribute, and right click an empty part of the canvas for Add date dimension, Add time dimension and Add currency dimension. Each change is a pending edit, and Undo steps back through them.
  4. Read the diagnostics. The Diagnostics tab on the right edge of the canvas counts the errors, warnings and info the derivation found and lists errors first. Each row names the object to act on, the cause and the remedy. A hub without a point-in-time table carries Create point-in-time, a missing role row carries Dimension and Fact buttons, and with the AI features enabled every row carries Explain.
  5. Choose Publish. Publish derives the full star and writes, in one save, the dimensions, the facts, a source view for every derived dimension and fact with the column mappings that load it, and your pending source metadata edits. A source view sits beside the primary source of its target. When the save succeeds, BimlFlex reports Publish succeeded and the strip closes.

Save is the other write on the command bar. It saves your pending source metadata edits, such as roles and business names, without publishing the star. With the AI features enabled, the strip also points to Draft with AI in the toolbar, which drafts the history policy, reviews measures and critiques the star. Every suggestion waits for your accept.

To hide the strip at any step, use its close button. Show guide brings it back.

Where to Go Next​