In data warehousing, two main approaches to data modeling are Tabular and Multidimensional Cube. Each method has its own strengths and weaknesses, and the decision on which to use depends on the needs of the business.

Tabular models are built on the relational database model and use the DAX (Data Analysis Expressions) language for calculations and data analysis. DAX is a formula language used to perform calculations and create calculated columns, calculated tables, and measure in tabular models. An example of a DAX calculation is:

1 = SUM(Sales[Amount])

Multidimensional cube models, on the other hand, use the MDX (Multidimensional Expressions) language for calculations and data analysis. MDX is a query language used to perform calculations and create calculated members in multidimensional cube models. An example of an MDX calculation is:

WITH MEMBER [Measures].[Profit Margin] AS ([Measures].[Profit] / [Measures].[Sales])
SELECT {[Measures].[Profit Margin]} ON COLUMNS, {[Product].[Product].Members} ON ROWS
FROM [Sales Cube]

In addition to calculated members, multidimensional cube models also support complex data analysis across multiple dimensions. This makes them ideal for organizations that require a hierarchical view of their data and complex data analysis, especially in financial applications where they are commonly used for financial cubes.

In conclusion, the choice between tabular and multidimensional cube models in data warehousing should be based on the specific needs of the business and the nature of the data. Organizations that require complex data analysis, including the use of calculated members, should consider using multidimensional cube models and the MDX language. On the other hand, tabular models and the DAX language may be a better option for organizations that have simpler data analysis needs.