DP-600 · Glossary
The terms, in one sentence each
140 terms from the DP-600 study guide. Each one links to the card that explains it.
A
- aggregate function also: aggregate functions, aggregation, aggregations
- A function that turns many rows into one value per group, such as SUM, AVG, COUNT, MIN or MAX. Study card: Aggregate data →
- autobinding also: autobind
- The automatic reconnection of a deployed item, such as a report, to the item it depends on, such as a semantic model, in the target stage. Study card: Deployment pipelines →
B
- Boolean filter argument also: Boolean filter arguments
- A simple condition on a column, such as Tickets[WaitMinutes] > 30, passed straight to CALCULATE instead of a table built with FILTER. Study card: Improve DAX performance →
- bridge table also: bridge tables, bridging table, bridging tables
- A table of key pairs placed between two dimension tables so that a many-to-many link becomes two one-to-many relationships. Study card: Relationships and bridges →
- Build permission
- A semantic model permission that lets a user create new content, such as reports, on the model, and also use Analyze in Excel, export and XMLA access. Study card: Reusable Power BI assets →
C
- cache.abf also: local data cache
- A file in the project's .pbi folder that holds a local copy of the model and its data from the last edit. The default .gitignore excludes it. Study card: PBIP projects →
- CALCULATE
- The DAX function that evaluates an expression in a filter context that you modify, by adding, replacing or removing filters. Study card: DAX variables and iterators →
- calculation group also: calculation groups
- A set of reusable DAX calculations, called calculation items, that you apply to any explicit measure instead of writing one measure per variation. Study card: Calculation groups and parameters →
- calculation item also: calculation items
- One reusable formula inside a calculation group, such as prior year, that reshapes whichever measure a report pairs it with. Study card: Calculation groups and parameters →
- cardinality
- The relationship property that says whether each related column holds unique values: one-to-many, one-to-one or many-to-many. Study card: Relationships and bridges →
- certification also: Certified
- An endorsement applied by an authorized reviewer to indicate that an item meets organizational quality standards. Study card: Endorse Fabric items →
- COALESCE
- A T-SQL function that returns the first argument that is not NULL, used to replace missing values with a default. Study card: Duplicates and NULLs →
- column-level security also: CLS
- A data access rule that hides selected table columns from members of a security role. Study card: Fine-grained access →
- composite model also: composite models
- A semantic model whose tables use more than one storage mode, or whose DirectQuery tables come from more than one source. Study card: Composite models →
- Copy activity also: Copy activities
- A pipeline activity that copies data from a source to a destination at any scale, handling column mapping and format conversion on the way. Study card: Ingest or access data →
- cross-filter direction also: cross filter direction, bi-directional, bi-directional filtering
- The relationship setting that decides which way filters flow: from one table only (single) or in both directions (bi-directional). Study card: Relationships and bridges →
- CTAS also: CREATE TABLE AS SELECT
- CREATE TABLE AS SELECT, a T-SQL statement that creates a new table and fills it with the result of a query in one step. Study card: Enrich data →
D
- data connection also: data connections
- A saved definition of where a data source lives and how Fabric signs in to it, which pipelines, dataflows and other items can reuse. Study card: Create a connection →
- data pipeline also: data pipelines
- A Data Factory item that chains activities such as Copy, Dataflow and Notebook into one workflow that runs on demand, on a schedule or on an event. Study card: Ingest or access data →
- Dataflow Gen2
- A low-code Data Factory item that uses Power Query to connect to sources, transform data and load it into destinations such as a lakehouse or a warehouse. Study card: Ingest or access data →
- DAX also: Data Analysis Expressions
- Data Analysis Expressions, the formula language of Power BI semantic models, used for measures, calculated columns, calculated tables and security rules. Study card: Semantic model star schema →
- DAX query also: DAX queries
- A statement that returns a table of data from a semantic model, starting with EVALUATE. Unlike a DAX formula, it creates nothing in the model or the report. Study card: DAX queries →
- DAX query view also: Write DAX queries
- The Power BI Desktop view, also offered as Write DAX queries in the service, where you write and run DAX queries against a semantic model. Study card: DAX queries →
- DAX variable also: DAX variables, VAR
- A named value, declared with VAR and used after RETURN, that DAX calculates once and then reuses without recalculating. Study card: DAX variables and iterators →
- DAX window function also: DAX window functions, windowing function, windowing functions
- A DAX function such as WINDOW, OFFSET or INDEX that returns rows by their position in an ordered, optionally partitioned set of rows. Study card: DAX variables and iterators →
- deduplication also: deduplicate, deduplicating, dedupe
- Removing repeated rows so that each entity appears once, either by dropping exact copies or by keeping one row per key. Study card: Duplicates and NULLs →
- default labeling also: default label
- A policy that gives an item a preset sensitivity label when the user does not choose one. Study card: Sensitivity labels →
- Delta table also: Delta tables, Delta Lake
- A table stored as Parquet files plus a transaction log, the open format Fabric uses for tables in OneLake. Study card: Choose a Fabric store →
- denormalization also: denormalize, denormalized, denormalizing
- Storing the same value redundantly, such as repeating a department name on every item row, so queries read one table instead of several. Study card: Denormalize for analytics →
- deployment pipeline also: deployment pipelines
- A Fabric lifecycle tool that copies items between workspaces assigned to stages, typically Development, Test and Production. Study card: Deployment pipelines →
- deployment rule also: deployment rules, data source rule, data source rules, parameter rule
- A setting on a target stage that swaps a data source, a parameter value or a notebook's default lakehouse each time items are deployed there. Study card: Deployment pipelines →
- derived column also: derived columns
- A column whose value is calculated from other columns, such as unit price multiplied by quantity. Study card: Enrich data →
- dimension table also: dimension tables
- A table that describes something you analyze by, such as product, customer or date. Its attributes are used to filter and group. Study card: Star schema →
- Direct Lake
- A Microsoft Fabric storage mode that loads columns of Delta tables from OneLake into memory when a query needs them, giving Import-like speed without a full copy at each refresh. Study card: Direct Lake framing and fallback →
- Direct Lake on OneLake
- A Direct Lake option that reads Delta tables straight from OneLake, can combine several Fabric items, uses OneLake security and has no DirectQuery fallback. Study card: Direct Lake options →
- Direct Lake on SQL analytics endpoint also: Direct Lake on SQL analytics endpoints, Direct Lake on SQL
- A Direct Lake option tied to a single Fabric data source through its SQL analytics endpoint, which handles discovery and security checks and allows DirectQuery fallback. Study card: Direct Lake options →
- DirectLakeBehavior
- A model property of Direct Lake on SQL analytics endpoints, with the values Automatic, DirectLakeOnly and DirectQueryOnly, that controls whether a query may fall back to DirectQuery. Study card: Direct Lake framing and fallback →
- DirectQuery
- A storage mode that keeps data at the source and sends a query to it each time a visual loads, so results stay current. Study card: Semantic model storage →
- DirectQuery fallback also: fallback
- When Direct Lake on SQL analytics endpoints cannot answer a query from memory, it runs the query as DirectQuery against the SQL analytics endpoint, usually more slowly. Study card: Direct Lake framing and fallback →
- downstream inheritance
- The automatic passing of a sensitivity label from an item to the items that depend on it. Study card: Sensitivity labels →
- downstream item also: downstream items, downstream dependency, downstream dependencies
- An item that gets its data from, or is built on, another item, such as a report built on a semantic model. Study card: Lineage and impact →
- Dual mode also: Dual storage mode, Dual table, Dual tables
- A storage mode in which a table answers from its in-memory cache or from the source, whichever the query needs. It suits dimension tables beside DirectQuery facts. Study card: Composite models →
- dynamic format string also: dynamic format strings
- A DAX expression that returns the format string for a measure, so the display format changes with context while the value stays a number. Study card: Calculation groups and parameters →
E
- endorsement also: endorsements
- A Fabric trust signal that marks an item as Promoted, Certified, or Master data. Study card: Endorse Fabric items →
- EVALUATE statement also: EVALUATE keyword
- The one required keyword of a DAX query. It is followed by a table expression, and the table that expression produces is the result of the query. Study card: DAX queries →
- Eventhouse also: eventhouses
- A Real-Time Intelligence item that holds one or more KQL databases for storing and analyzing streaming, time-based event data at scale. Study card: OneLake integration →
- eventstream also: eventstreams
- A Fabric item that ingests events from a source, can transform them on the way and routes them to destinations such as an eventhouse or a lakehouse. Study card: Catalog vs Real-Time hub →
- explicit conversion also: explicit conversions
- A data type conversion that you request yourself by calling the CAST or CONVERT function. Study card: Convert data types →
F
- fact table also: fact tables
- A table of measurements about events, such as sales lines, with numeric measures and a key for each related dimension. Study card: Star schema →
- field parameter also: field parameters
- A model feature that lets report readers pick, usually in a slicer, which measures or columns a visual shows. Study card: Calculation groups and parameters →
- filter context also: filter contexts
- The set of filters in force when a measure is evaluated: report filters and slicers, filters sent along relationships, and filters the formula adds. Study card: Semantic model star schema →
- formula engine and storage engine also: formula engine, storage engine
- The two parts that run a DAX query. The formula engine builds the query plan, and the storage engine returns the data and first aggregations from the model's in-memory store. Study card: Query and visual performance →
- framing also: framed, reframing
- The Direct Lake refresh step that reads the latest metadata of the Delta tables and points the model at their current Parquet files, without copying the data. Study card: Direct Lake framing and fallback →
- full outer join also: full join, full outer
- A join that returns every row from both tables and fills in NULL on the side that has no match. Study card: Join or merge data →
G
- Git integration
- The Fabric capability that synchronizes supported workspace item definitions with a connected Git repository and branch. Study card: Workspace Git control →
- Git status
- The synchronization state that shows whether a workspace item is synced, changed locally, changed remotely, or in conflict with Git. Study card: Workspace Git control →
- grain also: granularity
- What one row of a table stands for, such as one sales line or one day per product. Aggregating raises the grain. Study card: Aggregate data →
- GROUP BY
- The SELECT clause that divides the rows into groups, so aggregate functions such as SUM and COUNT return one row for each group. Study card: T-SQL analytical flow →
H
- HAVING clause
- The SELECT clause that filters groups after GROUP BY, by testing an aggregate value such as SUM(Weight) > 50000. Study card: T-SQL analytical flow →
I
- impact analysis also: impact analyses
- A Fabric feature that lists the downstream items and workspaces a change to an item might affect, and can notify their contacts. Study card: Lineage and impact →
- implicit conversion also: implicit conversions
- A data type conversion that SQL performs automatically, for example when a smallint is compared with an int, without CAST or CONVERT in your code. Study card: Convert data types →
- Import mode
- A storage mode that loads a snapshot of the source data into the model's in-memory engine, so visuals are fast but only as fresh as the last refresh. Study card: Semantic model storage →
- incremental refresh
- A refresh approach that partitions a large table by date and reloads only the most recent partitions, so refreshes are faster and use fewer resources. Study card: Incremental refresh →
- inline table-valued function also: table-valued function, table-valued functions, TVF
- A function that takes parameters and returns a table, defined by one SELECT statement. You query it like a table. Study card: Reusable T-SQL objects →
- inner join also: inner joins
- A join that returns only the rows that have a match in both tables and drops all other rows. Study card: Join or merge data →
- item pairing also: pairing, paired items
- The link between an item in one stage and its counterpart in the adjacent stage. A deployment overwrites paired items and copies unpaired ones as new. Study card: Deployment pipelines →
- item permission also: item permissions
- A permission granted directly on one Fabric item rather than through a role that applies to the whole workspace. Study card: Item permissions →
- iterator also: iterators, iterator function, iterator functions
- A DAX function, such as SUMX or AVERAGEX, that goes through a table row by row, evaluates an expression for each row, then combines the results. Study card: DAX variables and iterators →
K
- KQL also: Kusto Query Language
- Kusto Query Language: the read-only query language of Real-Time Intelligence, written as a pipeline of operators that pass a table from one step to the next. Study card: KQL query pipeline →
- KQL database also: KQL databases
- A database inside an eventhouse that stores tables of event data and is queried with Kusto Query Language. Study card: OneLake integration →
- KQL queryset also: KQL querysets
- A Fabric item that holds saved KQL queries and runs them against a KQL database, an Azure Data Explorer cluster or Azure Monitor data. Study card: KQL query pipeline →
L
- lakehouse also: lakehouses
- A Fabric item that stores structured and unstructured data as files and Delta tables in OneLake, built for Spark and data engineering. Study card: Choose a Fabric store →
- large semantic model storage format also: large storage format, large semantic model, large semantic models
- A semantic model setting that stores the model in a format whose size limit is the capacity size, not the default limit, and that helps XMLA writes. Study card: Large semantic models →
- least privilege also: least privileged
- The practice of giving a person only the access needed to do the job and nothing more. Study card: Workspace roles →
- left anti join also: left anti
- A join that returns only the rows of the first table that have no match in the second table. Power Query offers it as a join kind. Study card: Join or merge data →
- left outer join also: left join, left outer
- A join that returns every row of the first (left) table, with matching values from the second table or NULL when there is no match. Study card: Join or merge data →
- limited relationship also: limited relationships
- A relationship with no guaranteed one side, such as many-to-many or across source groups. It joins with inner join logic, so unmatched rows drop out. Study card: Composite models →
- lineage view also: View lineage
- A Fabric view that shows how the items of a workspace connect, plus their data sources and upstream items one level outside the workspace. Study card: Lineage and impact →
M
- mandatory labeling also: mandatory label
- A policy that stops a user from saving an item until a sensitivity label has been applied to it. Study card: Sensitivity labels →
- many-to-many relationship also: many-to-many relationships, many-to-many cardinality
- A relationship in which both related columns can hold duplicate values, so neither table is the one side. Study card: Relationships and bridges →
- master data
- Core business data, such as a list of stores or products, that an organization treats as the single agreed version for all its reports. Study card: Endorse Fabric items →
- measure also: measures
- A named calculation written in DAX, such as total sales, that is evaluated at query time for each cell of a visual and is never stored. Study card: Semantic model star schema →
- MERGE statement
- A T-SQL statement that compares a target table with a source and inserts, updates or deletes target rows depending on whether they match. Study card: Join or merge data →
- mirroring also: mirrored database, mirrored databases
- A Fabric feature that adds an external database or catalog to OneLake, either by continuously replicating its data as Delta tables or by referencing it in place. Study card: Ingest or access data →
N
- notebook also: notebooks
- A Fabric item where you write and run code, usually Spark with Python, Scala, SQL or R, to explore, transform and load data. Study card: Choose a Fabric store →
- NULL also: nulls, null value, null values
- A marker for a value that is unknown or missing. It is not zero or an empty string, and comparing it to anything returns unknown. Study card: Duplicates and NULLs →
O
- object-level security also: OLS
- A data access scope that grants access only to selected tables or folders in a data item. Study card: Fine-grained access →
- on-demand load also: on-demand loading
- A behavior of large semantic models that pages into memory only the data a query needs, so an evicted model becomes available again quickly. Study card: Large semantic models →
- on-premises data gateway also: on-premises data gateways, gateway cluster
- Software you install inside your own network that lets Fabric reach data sources there, such as a SQL Server instance that is not exposed to the internet. Study card: Create a connection →
- OneLake
- The single data lake that every Microsoft Fabric tenant gets automatically, where Fabric items keep their data in open formats such as Delta Parquet. Study card: Catalog vs Real-Time hub →
- OneLake availability
- A setting on a KQL database or table that keeps a read-only Delta Lake copy of its data in OneLake for other Fabric engines to use. Study card: OneLake integration →
- OneLake catalog
- The Fabric experience for finding and inspecting the items you can access, understanding their governance status and managing who can access them. Study card: Catalog vs Real-Time hub →
P
- PBIDS also: Power BI data source file, .pbids
- Power BI data source file, a small file that holds one connection definition and no credentials, and pre-fills Get Data for new report creators. Study card: Reusable Power BI assets →
- Performance Analyzer
- A Power BI pane that records how long each visual takes to load and splits the time into DAX query, direct query, visual display and other. Study card: Query and visual performance →
- Power BI project also: Power BI Desktop project, Power BI Desktop projects, Power BI projects, PBIP
- A Power BI Desktop save format that stores a report and its semantic model as folders of plain text files instead of a single .pbix file. Study card: PBIP projects →
- Power BI template also: Power BI templates, PBIT, .pbit
- A .pbit file that keeps a report's pages, data model definition and query definitions without any data, used as a starting point for new reports. Study card: Reusable Power BI assets →
- predicate also: predicates
- A condition that is true, false or unknown for each row, such as Region = 'West'. A filter keeps the rows for which it is true. Study card: Filter data →
Q
- QUALIFY clause also: QUALIFY
- A T-SQL clause that filters rows by the result of a window function, such as ROW_NUMBER, without a subquery. Study card: Filter data →
R
- RangeStart and RangeEnd also: RangeStart, RangeEnd
- Two reserved, case-sensitive Date/Time parameters in Power Query that mark the start and end of the date range loaded for each partition. Study card: Incremental refresh →
- Real-Time hub
- The tenant-wide place in Fabric to discover, connect and manage streaming data, including eventstream streams, KQL tables and Fabric events. Study card: Catalog vs Real-Time hub →
- reference table also: lookup table, reference tables, lookup tables
- A small table of descriptive values, such as region names, that you join to your data to add context. Study card: Enrich data →
- refresh policy also: refresh policies, incremental refresh policy
- The incremental refresh settings of a table: how much history to keep, how recent a period to reload each time, and optional settings such as detecting data changes. Study card: Incremental refresh →
- relationship also: relationships
- A model link between one column in a table and one column in another table, along which filters flow from one table to the other. Study card: Relationships and bridges →
- role-playing dimension also: role-playing dimensions
- A dimension, such as a date table, that relates to a fact table more than once so it can filter by different roles like order date and ship date. Study card: Semantic model star schema →
- row context also: row contexts
- The current row of a table while a formula runs row by row, as in a calculated column or inside an iterator. Study card: DAX variables and iterators →
- row-level security also: RLS
- A data access rule that limits a user to the rows that match the rule for a table. Study card: Fine-grained access →
S
- SELECTEDMEASURE
- A DAX function, usable only in calculation items and format string expressions, that stands in for the measure currently in context. Study card: Calculation groups and parameters →
- semantic model also: semantic models
- The model behind Power BI reports: tables, relationships and measures that reports, Excel and other tools query for numbers. Study card: Semantic model storage →
- semantic model eviction also: eviction
- A Premium behavior that removes inactive semantic models from memory so the capacity can load others, even when their total size exceeds its memory. Study card: Large semantic models →
- sensitivity label also: sensitivity labels
- A Microsoft Purview classification applied to content so an organization can identify and govern sensitive information. Study card: Sensitivity labels →
- sharing link also: sharing links
- A URL to a Fabric item whose audience can be your organization, specific people, or only people who already have access. Study card: Item permissions →
- shortcut also: shortcuts
- An object in OneLake that points to data in another location, so the data shows up as a folder without being copied. Study card: Ingest or access data →
- slowly changing dimension also: SCD, slowly changing dimensions
- A dimension that handles changing attributes. Type 1 overwrites the value, type 2 adds a new row version, type 3 keeps the previous value in a column. Study card: Star schema →
- snowflake dimension also: snowflake dimensions
- One business entity, such as product, spread over several normalized tables that point to each other in a chain. Study card: Semantic model star schema →
- snowflake schema also: snowflake schemas
- A variant of the star schema in which a dimension is normalized into several related tables, such as item, class and department. Study card: Denormalize for analytics →
- source group also: source groups
- In a composite model, the tables and relationships that come from one DirectQuery source, or from all imported sources together. Study card: Composite models →
- SQL analytics endpoint also: SQL analytics endpoints
- A read-only T-SQL query surface that Fabric creates automatically over the Delta tables of a lakehouse and other items. Study card: Choose a Fabric store →
- SQL query editor
- The text editor in the Fabric portal where you write and run T-SQL against a warehouse or SQL analytics endpoint, including DDL, DML and DCL statements. Study card: Visual query editor →
- SQL view also: SQL views, T-SQL view, T-SQL views
- A saved SELECT query that you read like a table. The data itself stays in the tables the query reads. Study card: Reusable T-SQL objects →
- SSMS also: SQL Server Management Studio
- SQL Server Management Studio, a client tool that can connect to a workspace as an Analysis Services server to run queries and TMSL scripts. Study card: XMLA model management →
- star schema also: star schemas
- A design with a central fact table surrounded by dimension tables, drawn like a star. It suits queries that filter, group and total. Study card: Star schema →
- storage mode also: storage modes
- A table property that decides where the table's data lives and when it is read: in memory, at the source, or in OneLake. Study card: Semantic model storage →
- stored procedure also: stored procedures
- A named block of T-SQL statements saved in the database and run on demand with EXEC. It can take parameters and change data. Study card: Reusable T-SQL objects →
- summarize operator
- The KQL operator that groups rows by the values in its by clause and computes aggregates such as count(), sum() or avg() for each group, returning one row per group. Study card: KQL query pipeline →
- SUMMARIZECOLUMNS
- A DAX function that returns a summary table. It groups by the columns you list and adds a column for each measure or expression you name. Study card: DAX queries →
- surrogate key also: surrogate keys
- A generated identifier, usually an integer, that a dimension table uses as its primary key instead of the key from the source system. Study card: Star schema →
T
- T-SQL also: Transact-SQL
- Transact-SQL, Microsoft's dialect of SQL. A Fabric warehouse and the SQL analytics endpoint of a lakehouse are queried with it. Study card: Reusable T-SQL objects →
- TMDL also: Tabular Model Definition Language
- Tabular Model Definition Language, a readable text format that saves a semantic model as a folder with one file per table, role and similar object. Study card: PBIP projects →
- TMSL also: Tabular Model Scripting Language
- Tabular Model Scripting Language, a JSON-based scripting format for creating, changing and refreshing the objects of a tabular model. Study card: XMLA model management →
- TRY_CAST also: TRY_CONVERT
- A T-SQL function that converts a value to another data type and returns NULL instead of an error when the conversion fails. Study card: Convert data types →
U
- user-defined aggregation also: user-defined aggregations, aggregation table, aggregation tables
- A summary table you map to a detail table, so queries at the summary level are answered from a small in-memory table instead of the large detail table. Study card: Query and visual performance →
V
- V-Order also: V-Ordered
- A Parquet file optimization that improves compression, so Direct Lake can load columns into memory and scan them faster. Study card: Direct Lake framing and fallback →
- VertiPaq
- The in-memory, column-based engine behind Import and Direct Lake tables, built to store columns compactly and to filter them quickly. Study card: Improve DAX performance →
- visual query editor also: visual query
- The no-code query tool in the Fabric portal where you drag tables onto a canvas and shape them with Power Query steps, and Fabric generates the T-SQL. Study card: Visual query editor →
W
- warehouse also: warehouses, Fabric Data Warehouse
- A Fabric item for relational, T-SQL based data warehousing on OneLake, with full DDL, DML and multi-table transaction support. Study card: Choose a Fabric store →
- WHERE clause
- The T-SQL clause that keeps only the rows that meet a condition. It runs before grouping and before window functions. Study card: Filter data →
- window function also: window functions
- A function used with OVER that computes a value for each row from a set of related rows, such as a rank within a group, without collapsing the rows. Study card: T-SQL analytical flow →
- workspace role also: workspace roles
- A role that grants a defined level of access to a Fabric workspace and all items inside that workspace. Study card: Workspace roles →
X
- XMLA endpoint also: XMLA, XMLA endpoints
- The Analysis Services protocol endpoint of a workspace that lets client tools query, script, deploy and refresh its semantic models. Study card: XMLA model management →