DP-600 Microsoft Fabric Analytics Engineer - Practice Test 1

Transform Data opens the Power Query Editor, which is used to connect to data sources, clean data, remove/rename columns, filter rows, and apply transformations before loading the data into the model. DAX measures are created in the Report or Data view, not here. See more: Power BI Desktop & Service

CALCULATE modifies the filter context. Using CALCULATE with ALL() removes existing filters from the specified table or column, allowing you to compute a total that ignores slicers and other report filters. SUM and SUMX respect the current filter context. TOTALYTD aggregates year-to-date, it does not remove filters. See more: DAX Functions

In a star schema, dimension tables (e.g., Date, Product, Customer) contain descriptive attributes that provide context for filtering and grouping. The fact table holds measurable, numeric event data (e.g., sales amounts, quantities). Dimension tables typically have far fewer rows than fact tables. See more: Semantic Models

Calculated columns are computed during data refresh and stored in the model - they operate row by row and increase model size. Measures are not stored; they are evaluated dynamically when a visual or query requests them, using the current filter context. Both can use CALCULATE. See more: DAX Functions

When you click Publish in Power BI Desktop, you select a workspace in the Power BI Service (Microsoft Fabric). The .pbix report and its embedded semantic model are uploaded to that workspace, making them accessible to users with appropriate permissions. See more: Power BI Desktop & Service

In a properly designed star schema, each dimension table has a one-to-many relationship with the fact table. The dimension side is "one" (unique key) and the fact side is "many" (foreign key values repeat). Filters flow from dimension to fact (single direction), enabling efficient slicing. Many-to-many should be avoided where possible. See more: Semantic Models

USERELATIONSHIP is used inside CALCULATE to temporarily activate an inactive relationship between two columns. This is common in date intelligence where a fact table has multiple date columns (e.g., OrderDate, ShipDate) but only one can be the active relationship to the Date dimension. RELATED returns lookup values; it doesn't manage relationships. See more: DAX Functions

A workspace in Power BI Service / Microsoft Fabric is a collaborative container where team members can publish, store, share, and manage items such as reports, dashboards, semantic models, lakehouses, and pipelines. Access is controlled through workspace roles (Admin, Member, Contributor, Viewer). See more: Microsoft Fabric Overview

ADDCOLUMNS takes a table and adds new calculated columns to it, evaluating an expression for each row. It returns a table. FILTER also returns a table but it is used to restrict rows, not add columns. CALCULATE modifies filter context and returns a scalar. SUM returns a scalar value. See more: DAX Functions

Row-Level Security (RLS) uses DAX filter expressions to limit which rows of data a user sees. For example, a sales rep sees only their region's data. RLS is defined in Power BI Desktop (Model view), published to the service, and users are assigned to roles. It does not control visuals or workspace membership. See more: Security, Governance & Deployment

DIVIDE(numerator, denominator [, alternateResult]) performs division and handles divide-by-zero gracefully by returning BLANK() (or an optional alternate result) instead of an error. Using the / operator in DAX would return an error on zero division. DIVIDE is the recommended approach for all division in DAX measures. See more: DAX Functions

Import mode loads and compresses data into the VertiPaq in-memory columnar store. This provides the fastest query performance but requires scheduled refreshes to keep data current. DirectQuery, by contrast, sends queries to the source at render time. Direct Lake is a Fabric-specific mode that reads delta tables directly without import. See more: Semantic Models

A dedicated Date table is required for time intelligence functions like TOTALYTD, SAMEPERIODLASTYEAR, and DATESINPERIOD to work correctly. It must have a contiguous date column marked as a "Date Table" in Power BI. Without a proper Date table, auto date/time tables are created automatically but they are less flexible. See more: DAX Functions

CROSSFILTER is used inside CALCULATE to change how a relationship filters in the current DAX expression. For example, you can change a one-way filter to bidirectional, or disable the filter direction entirely. This is useful when computing ratios or when you need filters to flow in both directions temporarily without changing the model's default relationship setting. See more: DAX Functions

Direct Lake is a Microsoft Fabric-specific storage mode that reads delta-parquet files directly from OneLake. It combines the performance of Import mode (in-memory VertiPaq engine) with the freshness of DirectQuery - the engine loads column segments on demand without a full data copy. It requires Fabric capacity (not available in shared Power BI capacity). See more: Microsoft Fabric Overview

OneLake is Microsoft Fabric's built-in data lake - a single, tenant-wide storage layer based on Azure Data Lake Storage Gen2. You don't need to create or configure storage accounts separately. All Fabric items (Lakehouses, Warehouses, etc.) store their data in OneLake. It supports the delta/parquet format and can be accessed via shortcuts from external sources. See more: Microsoft Fabric Overview

Dataflow Gen2 uses Power Query (M language) to connect to sources, apply transformations, and write the results to a supported output destination such as a Lakehouse, Warehouse, or Azure SQL. It is the evolution of Dataflow Gen1 and adds direct Fabric integration, faster refresh, and output destinations. It is not for real-time streaming. See more: Dataflows & Pipelines

HAVING filters rows after the GROUP BY aggregation has been applied. For example: SELECT Region, SUM(Sales) FROM Orders GROUP BY Region HAVING SUM(Sales) > 10000. WHERE filters individual rows before aggregation. ORDER BY sorts results. GROUP BY groups rows for aggregation. See more: SQL in Fabric

Microsoft Fabric Lakehouses store managed tables in the Delta Lake format, which uses Parquet files for the data along with a _delta_log transaction log folder. Delta Lake provides ACID transactions, schema enforcement, and time travel. CSV files can be placed in the Files section but must be converted to Delta tables to be fully queryable via SQL endpoint. See more: Lakehouse & Data Warehouse

In KQL, the `where` operator filters rows based on a Boolean predicate. Example: StormEvents | where State == "FLORIDA" | where EventType == "Tornado" This is equivalent to SQL WHERE. The `extend` operator adds calculated columns; `summarize` groups rows; `join` combines tables. See more: KQL & Eventhouse

A deployment pipeline in Fabric/Power BI provides a structured way to promote content through development -> test -> production workspaces. You can compare content between stages, deploy selectively, and set rules (e.g., different data source connections per stage). It is separate from data pipelines, which are used for ETL/ELT. See more: Security, Governance & Deployment

LEFT OUTER JOIN (or LEFT JOIN) returns all rows from the left table and the matched rows from the right table. When there is no match, NULLs appear for columns from the right table. INNER JOIN returns only matching rows in both tables. CROSS JOIN returns the Cartesian product (every combination). See more: SQL in Fabric

Sensitivity labels from Microsoft Purview Information Protection can be applied to Fabric items (reports, semantic models, lakehouses, etc.) to classify and protect data. Labels like Confidential or Highly Confidential can enforce encryption and warn users when exporting or sharing data. They propagate from semantic models to downstream reports automatically. See more: Security, Governance & Deployment

Each cell in a Power BI matrix visual creates a filter context from the intersection of all active filters: the row header value(s), column header value(s), slicer selections, report-level filters, and page-level filters. The measure is evaluated once per cell in this combined filter context. Understanding filter context propagation is fundamental to mastering DAX. See more: DAX Functions

Incremental refresh partitions a table by date and refreshes only the recent partitions during each refresh cycle. Historical data stays in older partitions and is not re-queried from the source. This dramatically reduces refresh duration and source load for large tables. It requires the data source to support query folding and uses RangeStart/RangeEnd Power Query parameters. See more: Semantic Models