drill_fields parameter controls what happens when a user clicks on the value of a table cell while Exploring data. When a user does this they “drill” into the data, allowing them to see the individual records that make up that cell, limit their query to the cell’s value, or slice the data in a related way.
drill_fields parameter accepts a list of fields, sets, or a combination of both fields and sets.
Please note that drilling can be disabled in some cases when the
can_filter parameter is used.
Drilling Into Dimensions
When you drill into a dimension, you have several options:
1. Limit the query to the dimension value clicked.
To do this, users click on a dimension value and choose to filter on it, as shown in this example:
This functionality always happens by default, and doesn’t require the use of
2. Limit the query to the dimension value you clicked, and replace the dimension with another.
Another option is to limit the query to a particular value, and replace the dimension with a different, related dimension. For example, you may have a country field, and want to be able to drill into it by state. The LookML for this would appear like:
In the Looker UI, this option will appear as:
In this example, if the user chooses to drill by state, the query will be limited to the USA (because that is the value they clicked on), and the Country dimension will be replaced by State:
Dimension groups of
type: time have drill fields added to them by default. Each timeframe can drill to the more granular timeframes (e.g. week can drill to date and time, but not month). However, you can limit the timeframes that can be drilled to by using the
Drilling Into Measures
Drilling into a measure shows the row level data about the items that comprise that measure. Consequently, while
drill_fields can be used on any measure type, it generally makes more sense if used with a
count_distinct measure. The information that is displayed for each row is defined by the fields or sets that you define in
For example, suppose you have the following LookML:
This will result in the following user experience: