Creating intervals of numerical data

You can group numerical data into intervals that are then handled as categories using binning.

About this task

The values in a numerical column might not be important individually. Sometimes bins of values are preferred because they can be shown as categories in a visualization. When you create bins, the range between the lowest value and the highest value in the numerical data column is divided into a number of intervals.

For example, if your data contains age information about a group of people, you might want to arrange the ages into a smaller number of age groups as illustrated below.
binning slider

You decide how many bins the range should be divided into. When binning the data, a slider is shown which can be dragged to show the desired number of bins. The bar chart below is based on the same data as above but adjusted to show more bins.
binning slider with many bins

Before you begin

A numerical data column is selected on an axis.

Procedure

Click the column selector on the axis to open its popover.
In the popover, select Auto-bin column.
The popover looks slightly different in the web client and in the installed client. In the web client, click , and then select Auto-bin column. In the installed client, you might need to click Expand before you see the option.

The range of values on the axis is divided into bins. A slider where you can change the number of bins is added next to the column selector.
Specify the number of bins by dragging the slider.

Tip: If you double-click the slider handle, you can enter the number of desired bins in a dialog. Then the range between the lowest value and the highest value in the numerical data column will be divided into a number of intervals of the same size.

Results

The visualization adjusts to reflect the specified number of bins. The endpoints of the intervals are automatically set to neat values.

Example

The table lists body heights and weights for 33 individuals, and each individual's height is represented in the sorted bar chart.

By coloring the bars by weight, and binning the numerical values, it is easy to distinguish individuals in certain weight intervals. In the image below, the range is divided into 4 bins. Individuals that might weigh too much for their heights, can be spotted.

For more examples, see Creating a histogram.