Okay, there’s a lot to take in here, but this exercise is one of the most integral to understanding Tableau’s data language. Take your time and refer back to the chart on the right as needed.

During the Loading Data lesson, we saw how Tableau assigns different field types to data (numeric, string, geographic, etc.). Tableau also categorizes each of these fields by their role in our visuals. Tableau calls the two roles *dimensions* and *measures*.

Dimension icons are blue and represent discrete fields: categories and descriptive variables.

Measure icons are green and represent continuous fields: numbers that we can perform calculations on.

Tableau is usually pretty accurate in assigning dimensions or measures to a data type, but it’s always good to double-check. Remember when we changed the `Borocode`

field from a number to a string? Tableau had assigned this field as a *measure* because it contained numbers. But we knew that the numbers represented different boroughs in New York City, and we changed the data type to a string so that Tableau will treat this variable as a dimension.

Why does this matter? Let’s walk through what happens if we let Tableau treat `Borocode`

as a measure.

When `Borocode`

is read as a Number and Measure, Tableau aggregates the `Borocode`

field in the `SUM()`

formula. (Notice that the `Borocode`

Pill is green and has been wrapped in the `SUM()`

formula.)

Since the field was “numerical” and set as a measure, Tableau summed the values in that field and created a bar graph to represent the totals. This doesn’t tell us anything meaningful because it doesn’t mean anything! Manhattan has just a small bar in the graph because its `Borocode`

is 1 rather than 4 or 5.

When we change the data type so that Tableau reads `Borocode`

as a String and Dimension, however, `Borocode`

now functions as an ID or label for each borough. Tableau will no longer attempt to do any calculations on this field, and we can see here how it corresponds with the variable `Boroname`

.

With that sorted out, we can use the field to make a map that uses the `Borocode`

variable to apply color to each borough, for example.

Let’s open our workbook with the Tree Census Data so we can explore another example together.

### Instructions

- Starting with the Tree Census data and a blank workbook, let’s look at the
`Tree Dbh`

field. It represents tree trunk diameter, measured in inches - that’s not obvious from the name, but rather something we can look up in the data dictionary. To make it clearer going forward, let’s rename it`Trunk Diameter`

. Right-click the field and select Rename to do this. - Drag the
`Trunk Diameter`

field to the Rows Shelf at the top of the window. Tableau will`SUM`

all the values that are available. It tells us that the sum of all the trees in the dataset is just under 8 million inches in diameter - a totally useless calculation. - Now pull the
`Borocode`

and`Boroname`

pills to the Columns Shelf, and we’ll see the sum of trunk diameters by borough. That’s more information, but this aggregation still doesn’t tell us whether Queens has*more*trees, trees with larger trunks, or both. So let’s compare the difference between measures and dimensions for the “Trunk Diameter” field. - Make a new sheet. Pull the
`Trunk Diameter`

field to the Rows Shelf again. - This time, click the arrow on the green pill to show the dropdown menu. Select
**Dimension**and the**Discrete**and*voila*, you’ll see all of the trunk diameters listed out one by one. Tableau now knows to show each of these values rather than summarizing them in the`SUM()`

formula. (We also dropped the`Tree ID`

field to the Filter pane and selected just a handful of trees - this is a great trick for working with large datasets, and we’ll cover it in more depth very soon.)