Data Granularity

Granularity refers to how grained the data is. Granularity concerns the detail level of the data or the size of data decomposition. Namely, granularity concerns the range of data considered as a whole during the analysis process.

The time range is common granularity. Assuming you are analyzing a customer's shopping data. The annual granularity concerns the customer's purchases for an entire year, the monthly granularity concerns the customer's purchases for each month, and the daily granularity concerns the customer's purchases every day. You can formulate different granularity according to data requirements, and the final observation values will change.

Table: The whole table represents the annual sales overview of the company.

Row: Each row represents the sales status for the entire year.

Column: Columns represent information such as total sales volume and total cost respectively.

Medium-Grained

Table: This table is more detailed and represents monthly sales data.

Row: Each row represents the monthly sales data.

Column: Columns represent information such as product category and sales volume respectively.

Fine-Grained

Table: This table is even more detailed and contains the specific information of each order.

Row: Each row represents a specific order.

Column: Columns represent information such as order ID, sales date, product name, and product quantity respectively.

In daily business scenarios, the choice of granularity depends on the user's data needs and analysis goals. Coarse-grained granularity is suitable for overviews and high-level analysis, medium-grained granularity provides more detailed information, and fine-grained granularity provides the most specific and detailed data.

In combination with product functions:

The Group Summary function in FineBI data edit can change data structure and granularity.