Once a dataset has been shaped and filtered, it probably still needs a good few modifications to make it ready for consumption. Many of these modifications are, at their heart, a selection of fairly simple yet necessary techniques that you apply to make the data cleaner and more standardized. I have chosen to group these approaches under the heading data cleansing.
The sort of things that you may be looking to do before finally loading source data into the data model can include the following:
Change the data type for a column—by telling Power BI Desktop that the column contains numbers, for example