Dimensions and Measures in Datasets
Dataset dimensions and measures are created when you import the data.
Data structures and data types are inferred during the initial data import, letting you immediately start working on story layout. For example, if you import this data:
Date | Name | Amount |
---|---|---|
2020/04/20 | Blueberry | 1 |
2020/04/20 | Grapefruit | 2 |
The resulting dataset could have a Date dimension, a String dimension, and an Integer measure, which you could start using in a story. If those inferred data types aren't what you want though, you can change them first. For example, if a column of dates is initially set to the String data type, you can change it to the Date type.
Columns of numbers are usually defined as measures, while other columns are defined as dimensions.
For measures, you can define the number of Decimal Places, and the Measure Units; for example, bottles or kilograms.
For dimensions, you can define one column to be the Description property for that dimension. For example, if you have these columns:
Product | Product Description |
---|---|
01-001 | Pants |
01-002 | Shirts |
01-003 | Dresses |
You could set the Product Description column to be the Description property for the Product dimension.
If you don't want to include all of the imported data columns as dimensions and measures in your dataset, you can delete them in the Dataset Overview panel, in the Output list (on a dimension token, select ). The raw data columns will still be included in the dataset, but dimensions and measures won't be created for them, and therefore won't be available to use in stories.
If you later decide that you do want one of those deleted columns to be made into a dimension or measure, select the column, and then in the Details panel, select Use as dimension or Use as measure.
You can easily create level hierarchies based on the imported dimensions. For details, see Creating a Level-Based Hierarchy in a Dataset.
For information about dimensions in models, see Dimensions in Models.