Tag Archive for: Power BI

Scenario: Management has requested an easy way to view a subset of strategic reports located in different Power BI workspaces. You can ask the users to mark reports and dashboards as favorites so they can access pertinent content in the Favorites menu, but you’re looking for an easier configuration, such as to create a book of reports with a built-in navigation that organizes reports in groups (like a table of contents), such as the screenshot below demonstrates.

Workaround: Creating a Power BI app might be your best option. However, a long-standing limitation of apps is that there is 1:1 relationship between an app and a workspace. Therefore, by default all included content must come from the same workspace and you can’t create multiple apps in a workspace, such as to distribute content to different audiences.

But thanks to the upgraded navigation experience, an app can include links to reports in other workspaces if consumers have at least Viewer permission to these workspaces.

1. Start by deploying the core set of reports that you want to distribute to a workspace, such as an Executive Reports workspace. This workspace will serve as a base for the app.
2. Create an app. In the Navigation tab, the app will already include sections for each distributed report. Add links to reports in other workspaces and organize them in group.
3. To prevent external reports “losing” the navigation experience, specify the report embed URL which you can obtain by opening the report in Power BI Service, and clicking the Embed , “Website or portal” menu. Then, copy the URL (not the iframe element) and use it as the link.
4. Grant permission to the appropriate users or groups.
5. (Optional) Upload an image for the app logo so end users can tell the app apart from other apps.

Here is a configuration for a report residing in a different workspace. Notice that to preserve the navigation experience, the report is configured to open in the page content.

There are two main caveats of this approach:

1. Users must have permissions to the workspaces (except the app workspace) where the distributed reports reside. The app permissions you set up don’t propagate to content outside the app workspace.
2. App consumers are not isolated from changes to the external reports. By default, an app propagates content changes to included content only when the app is updated. End users will see changes to external content even if the app is not updated.

Besides the built-in cross-filtering and cross-highlighting among visuals, Power BI supports two explicit filtering options: slicers and filters. Which one to use? Traditionally, you would use a slicer when you want the user to easily see what’s filtered on the report page. But with the introduction of the new filter pane and slicer enhancements, the choice becomes more difficult. Let’s compare the two options:

A glaring gap for both filters and slicers is that you can’t currently set the default value programmatically, such as to default a date filter to the last date with data. As a workaround, you can add a field to the Date table, such as IsToday, that is set to Yes for the last date and prefilter on this field, but users must be educated to know how to remove the filter if they want to select another date. This is especially cumbersome with slicers, which don’t even support a single date selection, unless configured as a drop-down or Before/After.

Based on experience, people tend to rely mostly on slicers. But because it’s not uncommon to create reports that must be visually appealing on desktops and mobile devices, here are some recommendations to accommodate both large and small displays.

1. Use slicers judiciously because they take space on the report page. This could be an issue with mobile devices. Besides taking space, mobile users find it difficult to select values in slicers. I typically use slicers only for common filters, such as Date. Another scenario where slicers could be useful is when you need to visualize the items in a special way, such as a slider or to show a hierarchy of items.
2. Use filters for the rest of the filtering needs, especially if you plan to optimize reports for mobile viewing in portrait mode and/or use Power BI Embedded to embed reports.

Power BI incremental refresh (a Power BI Premium feature) refreshes a subset of a table with imported data. The main goal is to reduce the refresh time so that new data becomes available online faster. Patrick LeBlanc has a great video about how to make the incremental refresh even more incremental by using the “Detect data changes” feature and he explains in detail how it works.

What if you want to fully refresh the dataset set up for incremental refresh? For example, you configure a table for incremental refresh periodically, but you want to fully process the dataset nightly, such as to pick the latest changes to dimensions. Currently, the only option to fully refresh the dataset with an incremental refresh policy is to republish the dataset and refresh it (this works because the first refresh is always full). When the XMLA endpoint becomes writeable, you’ll have the option to do so in an XMLA script. For example, the following script fully refreshes the InternetSales table without applying the refresh policy settings. Notice also the effectiveDate setting that allows you to overwrite the current date for testing purposes.

{

“refresh”: {

“type”: “full“,

“applyRefreshPolicy“: false,

“effectiveDate”: “10/24/2019”,

“objects”: [

{

“database”: “AdventureWorks”,

“table”: “InternetSales”

} ] } }