Power BI - dataflows

07-06-2019  0 Comment(s)

According to Microsoft this new addition is for self-service ETL by business annalists and BI professionals. A minor addition from our side: especially business annalists that are skilled with tools like Excel and BI professionals with already a focus on Power BI will probably be very exiting using this new addition. An other group that could benefit of this new feature is the technical application administrator that has a great amount of knowledge of the data model of their application. With dataflows the could do some data preparation for the business annalists

First of all you must have at least a Power BI Pro license and this new preview feature is a Power BI Service-only feature that is not (yet?) available in Power BI Desktop. However you can use the result of a Data Flow which we will show you in the last step.

1) Power BI Service
To create a dataflow, sign in to Power BI Service and go to one of your workspaces, but not My Workspace. Or create a new workspace. Here you will find a new tab called "Dataflows (preview)" and in the Create-menu there is a dataflow option. Click on it to create a new dataflow.


2) Add new entities
In this step we will add a source for our dataflow. You can map this data to one of the standard Common Data Model (CDM) entities, but you can also create 'custom' entities that are not mapped. The CDM is a standard model for example for contacts or accounts to which you can map your sources like CRM or SalesForce. It should make it easier for development, but also analytics. In a subsequent post we will explain the CDM.

For this example we will not map to CDM, but create a 'custom' entity. We will use a text file that contains sensor data. We use Azure Blob Storage to store this data, but of course you can also use an internal file share. However then you first need to install the On-premises data gateway.

As told, we choose "Azure Blobs" as data source and next fill in the connection settings. If you have ever used this storage account before then it will remember the account key. Choose your Blob folder which contains the data and now you can transform your data similar to the Power Query Editor in Power BI Desktop.

Comment Here


No Comments to Show