Tableau Fundamentals: Dimension vs. Measure

Comentários · 31 Visualizações

This tableau certification course will help you build interactive dashboards and master data visualization. This eLearning training provides a 100% money-back guarantee. Enroll now!

The third thing that we would have known the very first day we started using Tableau is that whenever connecting to data Tableau will categorize each data field into a measure or measure. Tableau will then categorize the fields based on their measure or dimension classification on the left portion of your workspace.

Knowing the distinctions between measurements and dimensions will make it easier to use the data available in Tableau.

What is a measurement?

As per the Tableau Knowledge Base, a measure is an area that is dependent which means that the dimensions of one or two determine its value.

Tableau uses any field that contains numbers (quantitative) data as an measurement.

Look at the following bar chart that was created in Tableau using the Sales measure from the Sample Superstore data set: Superstore Data set

Sales is a quantitative field, and it is expected that Tableau is able to determine that it is an indicator. It is as a dependent variable because an individual measure does not offer much value.

The amount of $8,951,931 is not meaningful by itself. It's dependent on context, which takes the form of being broken down into dimensions.

What is an aspect?

According to the Tableau Knowledge Base, a dimension is a term used to describe a field that can be considered an independent variable. Tableau treats any field with categorical or qualitative data as an element as a default feature.

Here's the exact sales measure that was mentioned the previous paragraph, split into the region dimension:

Since our sales figure is broken down by regions, we're now in a position to begin gaining insight from the information.

One of the insights that emerges from this is that the South region is relatively poor in sales compared to other regions.

This is a description that was realized only when we put measurements and dimensions.

Generallyspeaking, the measurement is the number. The dimension is the one it is that you "slice and dice" the number.

However, there may be some exceptions to this Therefore, it is helpful to know the way Tableau deals with these types of fields.

Think about a company that has distinct order IDs with numbers for every sale (i.e. Order 1 gets one number; and order two is given the same number etc.).

In the above definition, Tableau will classify this Order ID field as a measure the first time you connect to a data set with the field. However, the Order ID can be viewed as a measure due to the fact that we would "slice and dice' measures such as sales, by the Order ID to find out how much the revenue we earned per order.

Another rule we use is that if it's not making sense to combine an amount, it's likely to be a dimension.

It's exactly the same for the possible Order ID field just mentioned. There's no benefit in adding each of the Order ID numbers to get the total. And sure that this particular field ought to be considered a dimension, not measuring.

Another instance which comes to my mind when Tableau could mistakenly classify fields is if it comes to a particular field, which should be a measurement, but includes the word NULL within the first entry in the header of your column of your database. The word NULL is considered a string to Tableau and therefore qualitative, which will cause Tableau to categorize the field as an element.

Tableau Training can help you master the skill in the most comprehensive manner!

The positive side is the fact that any area not properly classified can be changed by right-clicking the field inside the Dimensions Shelf or Measures Shelf and choosing "Convert to dimension" or "Convert to measure" depending on the appropriate.

The same effect is possible by dropping the field on the Dimensions Shelf or Measures Shelf.

A thorough understanding of how measures and dimensions function in Tableau and the basics of data preparation mentioned (when appropriate) will aid in the creation of visualizations in the future.