Siirry pääsisältöön

Use Dataflows




Use dataflows to organize your work and to separate ETL from visualizations. Trust me, for I'm a stranger you found in a blog, this will change your report designing habits for good.

Your lineage should go from data source(s) to PBI report/dashboard/app like this:

    DS --> Dataflow -> Report

Datasources can be wherever they might be as long as you can connect to them from the PBI Desktop and PBI Service.

Dataflow will be designed in the Power Query Editor of  PBI Desktop and in the Model and Data views of PBI Desktop and published in PBI Service.
Report will be designed in the Power Query Editor of PBI Desktop and in the Model, Data and Report views of PBI Desktop and published in PBI Service.

The above stunning graph in more verbose format:

  1. Connect to your data source(s).
  2. Get the raw'ish data.
  3. Do the thing in the Power Query. Do the cleaning, transformations etc. Make the data so good that there's nothing for the usual end user to tinker.
  4. Disable all the loads except for the ones you wish to be seen by the dataflow users.
  5. Create a data model using the Data view of PBI Desktop. Make it nice. Like nice.
  6. If there's common calculations or measures to be done, do them. Use PQ or DAX, depending on the needs and your skill level.
  7. Keep away from the Report view. Hush. There's no need to go there. Forbidden. You remember Adam and Eve? This is your apple. Don't touch it or you'll be banished from the garden of low hanging fruits called clean dataflow.
  8. Publish your dataflow to PBI Service. Give it a descriptive yet good, sturdy name. Have a small party of 5 people with non alcoholic drinks and light snacks. Be considerate and take different allergies into account. Being a vegan is not an allergy, so you can freely not invite those kind of people.
  9. Design your report in PBI Desktop. Use the freshly created dataflow as a data source. Take some time to cherish the fact that there's no need to tinker with data as usually. Just hop on to designing the most fabulous report ever done in the history of fabulous reports. No more twiddling with text inputs in date fields! No more fiddling with data models! Just go on and create. Be the Bob Ross of PBI. There are no mistakes, no happy accidents, just pure enjoyment of clean, fresh data.

  10. Publish your report.
Bonus step: If it turns out (as it should) that the dataflow is pure gold and source of the ultimate truth, promote it or even certify it, if your tenant allows it.

Tunnisteet:

Kommentit

Lähetä kommentti

Tämän blogin suosituimmat tekstit

Understanding Power BI Context: What You Need to Know

It's crucial to comprehend the concept of context in order to use Power BI efficiently. In this article, I'll examine what context in Power BI is and how it can be utilized to produce reports that are more useful. What Are the Power BI Contexts? Row context and filter context are the two different types of context available in Power BI. These contexts specify how a report's data is filtered and computed. What Distinguishes Row Context from Filter Context? Whereas row context applies filters to specific data rows, filter context applies filters to the entire report. In other words, row context filters the data after it has been calculated, whereas filter context filters the data before it. What Does Power BI's Context Filter Mean? A filter type known as a context filter modifies the report's filter context. As a result, you can filter report data without altering the underlying data. You could add a context filter to a report that displays sales by region to only dis
Lähetä kommentti
Lue lisää