Hello everyone, welcome to my first blog. Well, my topic of interest is Tableau, and being an explorer of Tableau when I was digging into the concepts of Tableau, I landed on the very basics. Fine, that is how I decided to start writing about Dimensions and Measures.
When a data source is connected to the tableau, it assigns roles to the data fields based on the type of data the field contains. Data fields with textual values will be categorized as Dimensions and data fields with numerical values will be categorized as Measures. Both can be discrete and continuous. Discrete Dimension and Continuous Measure are more common.
Below are the dimensions and measures assigned by tableau by default, when connected superstore sample data source.
Dimensions contains “Qualitative values” which affects the level of detail in the view. It can also be called as independent variable. Dimensions dragged into the view increase the marks in the view.
Measures contains numeric “Quantitative values” which can be aggregated by tableau by default when we drag the measure into the view. Measures can also be called as dependent variables. Because the measure alone in the view has no meaning unless it is connected to dimension.
Let’s identify the role of the fields with their colors:
When a field is dragged into the view, the tableau gives green color for the continuous fields and blue color for the discrete fields.
The layout of the view depends on the fields whether they are discrete or continuous. The continuous field makes the axis to the view and the discrete field forms a header to the view.
Green Measure - When we drag the profit into the rows, it will be aggregated into SUM(profit) and will be displayed in green color.
Blue Measure - There is an option in tableau to change the role of the data field. Let’s change the Sum (Profit) as discrete from the dropdown option.
Blue Dimension - By default, the Order date field is discrete, when dragged into rows it forms the header for the view.
Green Dimension - In cases when we need to aggregate the date field, we have to change the role into continuous. Let’s change the YEAR (Order Date) to continuous from the dropdown option.
Note: Dimension fields with data type ‘string’ or ‘Boolean’ can not be changed into continuous.
Difference between Discrete and Continuous values when we use color in Bar chart visualization:
We use colors in our layout to better visualize the result.
When we drag and drop the discrete fields into colors under “Marks”, different colors will be given for discrete variables of the dimension. When drag continuous fields into colors, a gradient color will be given.
For instance, let us take category field from the superstore data source which has three variables Furniture, Office Supplies and Technology.
Let us investigate the sales for different categories by dragging Sales into columns.
If the Category is dragged inside color under “Marks”, different colors will be given for different categories.
If Sales is dragged inside color under “Marks” gradient color will be given.
Difference between discrete and continuous values when we use colors in a Map chart.
As we have seen in the case of the bar chart, in the map chart also, continuous values will be given a gradient color. Added to that, when we choose a continuous field to drag and drop into color under “Marks”, it will the map with the gradient color.
But when we drag and drop a discrete field into color under “Marks”, it will only create colored symbols. The map will not be filled with color. Let’s see an example below.
Measure Names and Measure Values:
When a data source is connected to Tableau, it creates five auto-generated fields. Those fields will be in italic font in the data pane. Measure Names is one of the autogenerated fields which along with another autogenerated field, Measure values, gives all the measures in the data pane with their aggregated values.
Let’s understand more with the example.
When we look at the data pane, the autogenerated fields are in italic font. They are Measure Names, Latitude(generated), Longitude(generated), Orders (Count), and Measure Values. Of these let’s investigate Measure Names and Measure values.
Let’s drag Measure Names into columns and Measure Values into rows. As a result, all the measures (Discount, Quantity, Profit, and Sales) from the data source and one autogenerated measure, (count of orders) are displayed. Measure values into the label display the numerical value on top of each bar. They are the total count for each measure.
I hope this blog is useful. Thank you