PL-300 key terms

The window inside Power BI Desktop where you shape data before loading it: choosing sources, removing columns, changing types, and combining tables. Each action you take is recorded as a step you can revisit or delete.

The ordered list, shown in the Query Settings pane, of every transformation a query performs. You can rename, reorder, edit, or delete a step, and Power Query re-runs the whole list each time the query refreshes.

The formula language behind Power Query, shown in the Advanced Editor and the formula bar. Every button you click in the editor writes M, and you can hand-write it for logic the interface does not offer.

A behavior where Power Query translates your transformations into a single query that the source database runs, so less data is downloaded. If a step breaks folding, later steps are processed locally instead.

A Power Query feature that summarizes a column's values, showing how many are distinct, unique, or empty and how they are distributed, so you can spot quality problems before you load the data.

Joining two queries into one by matching rows on one or more key columns, similar to a SQL join. The result can keep all rows from the first table, both tables, or only matching rows.

Stacking two or more queries that share the same column structure into a single table, adding their rows on top of each other. It is used to combine monthly files or regional extracts.

A Power Query transformation that turns wide, cross-tab style columns into attribute-value row pairs. It is often used to normalize data where month or category names appear as column headers.

A Power Query transformation that spreads the values of one column into new columns, turning row-based data into a cross-tabulated layout. It is the reverse of unpivoting and is used less often in models.

The classification assigned to each column, such as whole number, decimal, date, or text, that determines which operations and aggregations are valid. Wrong types cause broken relationships and incorrect results.

A query setting that keeps a query available for other queries to reference while preventing its rows from being loaded into the model. It is used for staging queries that hold intermediate transformation work.

A model design with one central fact table of measurable events connected to surrounding dimension tables that describe those events. It keeps relationships simple and is the layout Power BI performs best with.

The table in a star schema that stores transactional rows and numeric measures, such as sales amounts or quantities, plus foreign keys to dimensions. It is usually the longest table and the target of most aggregations.

A table that holds descriptive attributes used to slice and group facts, such as product names, customer cities, or calendar periods. Dimensions are typically wide and short, with a unique key per row.

The rule describing how rows in two related tables match, such as one-to-many or one-to-one. It determines which side is unique and affects how filters flow and how DAX evaluates relationships.

A relationship property that sets whether filters flow from the one side to the many side (single) or in both directions. Bi-directional filtering can create ambiguous paths and is used sparingly.

A column added to a table using a DAX formula that is evaluated row by row and stored in the model. It increases model size and recalculates during refresh, so prefer Power Query when possible.

A DAX formula that is calculated when a visual queries the model, based on the filter context at that moment. Measures are not stored as data and are the recommended way to define aggregations.

Data Analysis Expressions, the formula language used for measures, calculated columns, and calculated tables in Power BI. It works with tables and columns and behaves differently depending on filter and row context.

The set of filters active when a measure is evaluated, coming from slicers, visual rows and columns, report filters, and relationship propagation. Changing the filter context is how CALCULATE modifies a result.

A calendar table with one continuous row per date, marked as a date table, that time intelligence functions rely on. It must contain unique, gap-free dates and relate to the fact table's date keys.

The setting that determines whether a table's data is cached in the model (Import), queried live at the source (DirectQuery), or both (Dual). It affects refresh, performance, and which features are available.

The level a report filter applies to: one visual, one page, or all pages. Visual filters are narrowest and report filters are broadest, and all active levels combine when results are calculated.

A report setting that controls how a selected visual filters the other visuals on the page. You can make a target visual filter, highlight, or remain unaffected by the selection.

A feature that lets a user right-click a data point and jump to a separate page filtered to that item, such as a customer or product detail page. Target pages show contextual values and offer a back button.

A saved snapshot of a report page's filter, slicer, and visibility state. Bookmarks power navigation buttons, storytelling sequences, and toggles that switch between different views of the same page.

A hidden report page designed to appear when a user hovers over a visual, showing extra charts or details about the hovered data point. It replaces the default small tooltip with a custom layout.

A setting that links slicers across multiple pages so a selection made on one page carries to the others. You choose which pages each slicer controls and whether it stays visible there.

Formatting that changes colors, icons, or data bars based on rules or field values instead of being fixed. It draws attention to outliers and thresholds in tables, matrices, and charts.

A built-in visual that lets users break an aggregated value down by successive dimensions, choosing the next field at each level. It supports both manual expansion and AI-driven high-value expansion.

The panel in Power BI Desktop used to add reference lines, trend lines, forecasts, and error bars to a visual without writing DAX. Availability depends on the visual type and the model's data.

A shared container in the Power BI service where teams store and collaborate on reports, datasets, dashboards, and dataflows. Access is controlled by assigning roles such as Admin, Member, Contributor, or Viewer.

A packaged, read-only way to distribute a group of related reports and dashboards to an audience. Users receive the app rather than direct workspace access, and updates are published as new versions.

A model feature that restricts which rows each user can see, defined with DAX filters on roles and configured in Desktop. It is enforced in the service for viewers and should be tested with View as role.

A dataset permission that lets a user create new content, such as reports or composite models, on top of an existing dataset without editing the dataset itself. It is granted separately from workspace roles.

A Microsoft Purview label applied to reports, datasets, or dashboards that classifies the data and can enforce encryption or restrictions when the item is exported or shared. The label stays with the item.

A promotion (Promoted) or certification (Certified) badge applied to content so users can identify trusted, validated reports and datasets. Certification requires permission and usually passes through a review process.

A service feature that moves content through staged environments, typically development, test, and production, allowing comparison and selective deployment between stages. It requires capacity and assigned permissions.

A service component that lets the Power BI service reach data sources that are not publicly accessible, such as on-premises databases. It handles scheduled refresh and DirectQuery traffic between cloud and source.