The DAX data modeling engine is made up of six key components.
Importing data is the initial stage in creating a data model. There are numerous data sources that can be used, and after being imported, those data sources will be stored in two-dimensional tables. It is possible to use sources that are not two-dimensional, however, these sources must be transformed into a two-dimensional format either before or during import. A wide variety of functions are available in the query editor to aid with this kind of modification.

The Data View, the second component, allows you to view the unprocessed data for each table in the model. Clicking a table name from the list of tables in the right-hand panel activates Data View, which can display data for one table at a time. Although columns can be sorted in this view, doing so has no bearing on any visual sorting on the report canvas. Using the Report View, columns can be renamed, formatted, deleted, hidden, or have their datatype defined. A hidden column will never be shown in any field list in the report but will always be visible in the Report View. Both Report and Data View have the option to add or modify computations.

Tables
Tables are used for data management and archiving. Source data or the results of DAX calculations are the components of columns in tables.

Columns
There may be one or more columns in each table. Information from a single column is stored by the underlying data engine in a distinct index. In contrast to T-SQL, DAX stores data in columns rather than rows. Data is viewed as static and cannot be changed after it is loaded into a column. Columns can be referred to as fields.

Relationships
Two tables can be connected using a relationship that is described in the model. The only field in each table is used to describe the relationship. Only one-to-one and one-to-many relationships are possible. Multiple persons cannot build relationships with one another. Relationships are most typically used in DAX to offer to filter, rather than imitate data normalization for OLTP operations.

Measures
An individual result is returned by a DAX calculation known as a measure, which can be used in report visualizations or calculations for additional measures. A measure might be as simple as the total number of columns or rows in a table. Mechanisms of response and reaction to user interaction

Hierarchies
Hierarchies enable users to drill up and down through collections of two or more columns using interactive charts and visuals. An example of a typical hierarchy could be overdue data that creates a three-level hierarchy over the year, month, and day.
Other common hierarchies may be constructed using structures that represent organizational groupings in HR or Product data, or they may be based on geographic information (country, city, or suburb).

Learn a step-by-step guide on the DAX data modeling engine from our Youtube Video on Power Bi Training!