Data modeling and table properties

Data Modeling in Power BI

Maarten Van den Broeck

Content Developer at DataCamp

What is a data model?

Conceptual view of data elements
Typically a visual representation
Data models include:
- Tables
- Columns
- Relationships between tables
- Data types
- Keys

general view of a data model

Data modeling

The process of creating a data model
Why model data?
- Data $\neq$ perfect
- Reshape data for analysis
- Compress data usage
- Easier to understand model

data modeling

Data modeling in Power BI & Power Query

Power Query is the data preparation tool of different Microsoft products, including Power BI
Main goals:
- Manage queries
- Data modeling
Data modeling: 80% in Power Query, 20% in Power BI

Data modeling in Power Query

Columns and row management

Operation

Keep or remove specific columns
Keep or remove specific rows
Split a single column in multiple columns
Summarize/group rows in a table by the contents of a column

Example

Remove empty column
Keep top row as header
DD/MM/YYYY column split in DD, MM, YYYY columns
Sum or median of all rows

Data types

Choosing the right data type is essential:
- Constrain data to a specific shape
- Optimize storage
- Enable specific functionality
Power Query infers data type on first few hundred rows

Data types in Power BI

Rounding

Power Query

Actually changes the data, not just formatting
Typically not the right answer

Power BI

Changes how the data appears, not how it's stored
Generally a better answer than rounding in Power Query

The dataset

United States Census Bureau survey data of manufacturers
Summary statistics for manufacturing firms
North American Industry Classification System (NAICS)

Let's practice!

Data Modeling in Power BI

Preparing Video For Download...