Creating visualizations and graphs

Stay organized with collections Save and categorize content based on your preferences.

This page explains how to create graphics and charts, based on the results of a query, to best showcase your data. Looker keeps your query details and visualization configuration data together. When you share a query, recipients get your visualization as well as the data.

Quick guide

You can add an eye-catching visualization to any query result set on an Explore.

  1. Create and run your query.
  2. Click the Visualization tab.
  3. Select the type of visualization that best displays your data. For more options, click to the right of the displayed visualization options.
  4. Click Edit to configure the visualization option settings, such as naming and arranging chart axes, choosing the position and type of each data series, or modifying the chart color palette.

You can further customize your visualization by specifying which dimensions and measures to include in the visualization. If your data is missing key values, you can tell Looker to fill in those values on the appropriate part of your visualization.

Choosing a visualization type

After you create and run your query, click the Visualization tab in the Explore to configure your visualization options. Use the chart buttons to pick a visualization type.

The visualization type that you select determines how Looker represents the data series in your chart. A data series is a set of related data points plotted on a chart. For example, the number of orders placed each day for a set of dates is a series. In a column chart, a series is represented by columns of the same color; in a line chart, a series is represented by a single line. You can see a list of the series for your chart in the series menu, and on the chart legend.

Customizing visualizations with chart settings

You can customize a visualization to make your data more readable and to add visual styling. Click Edit to see the visualization options, then change the settings to get a result that suits you.

To see the visualization options available for a particular visualization type, click that type on the Visualization types documentation page.

This example shows some of the visualization settings for an area chart with stacked series.

Including multiple visualization types on a single chart

You can create charts that include more than one visualization type:

  1. Click the Edit button to show the customization options.
  2. Click the Series tab.
  3. In the Customizations section, an entry appears for each series in the chart. Click the arrow next to the series to display its customization options.
  4. In the Type box, select the type of visualization to use for that series.

Charts with multiple series types always layer line series and scatter series in front of area, column, and bar series.

To alter the layering order of column, bar, and area series, change the series' positions in the data table and click the Run button. The leftmost series layers on top and the rightmost series layers on bottom.

Creating stacked charts with multiple visualization types

You can include stacked series in a chart with multiple visualization types. All series of the same type as the chart overall will stack together; series of other types will not stack. For example, the chart below is a column chart, so the columns stack, but the line series do not stack.

To create a stacked chart that uses multiple y-axes, drag any series to a different axis in the Y menu. The stacked series will appear together. All other series can be moved independently, including individual series within a pivot.

Specifying LookML fields to include in the visualization

Looker adds all dimensions and measures that are selected in the field picker to any visualization. Sometimes, you might not want to display every dimension or measure in the chart.

Hiding fields from visualizations

Looker does not re-run queries to exclude fields or values that are hidden using Hide from Visualization or Hide "No"s from Visualization for table calculations. As a result, calculations based on fields with hidden Explore values may display unexpected results.

To hide a field from the visualization:

  1. Select the gear icon at the top right corner of the column.
  2. Select Hide this field from visualization.

In this example, the LookML field Accidents Total Fatalities is hidden from the visualization, leaving only Accidents Serious Accidents Count and Accidents Fatal Accidents Count in the chart.

To enable or disable a charted series, click that series in the visualization's legend. When disabled, the series color turns gray in the legend and the data disappears in the chart. Click the series again to re-enable it.

To hide table calculations from a visualization, see the instructions on the Using Table Calculations page.

Filling in missing dates and values

Some datasets have values, such as dates, that follow a predictable pattern. You might pull data by a timeframe and find that some dates, weeks, months, or other date types don't have a corresponding value. By default, the data table and the visualization will display the dates that the query returns and skip any missing dates. Looker's dimension fill option lets you display the missing dates or other values in the data table and on the corresponding axis of the query's visualization.

For example, this accident data from 1990 shows only a few dates, those on which an accident occurred:

If you do not use dimension fill, Looker connects the data points it has, resulting in a potentially misleading graph:

Turning on dimension fill adds the missing dates and makes the graph more informative:

To use dimension fill, select the appropriate dimension's gear menu in the Data section of an Explore. Choose the Fill in Missing Dates or Fill in Missing Values option:

Dimension fill is available for dimensions with yes/no values, tiered values, and most date types. It can also be applied to any dimension based on a list of values, via the case or tier parameters.

Dimension fill turns on automatically for queries that run with a single dimension and/or a single pivot, as long as you haven't applied filters to any measures. Dimension fill can also be applied to multiple dimensions at once in a query — including pivoted dimensions — however, Looker may automatically disable dimension fill to optimize query performance if it detects that too many fields will be generated with filled values.

You cannot use dimension fill in several cases:

  • When your Looker developer used the order_by_field parameter or disabled the allow_fill parameter on certain dimensions.

  • When dimensions have a filter applied to them and also have a fixed number of values, such as yes/no, days of the week, days of the month, and so on. Filtering against these field types eliminates the values that Looker needs to predictably and accurately fill in any missing values.

  • When you are drilling into a pivoted dimension.

  • When Looker detects that too many rows or columns will be generated with filled values and automatically disables dimension fill to optimize query performance.