Filtering records

In Wrangler, you can filter records for all types of columns, except Boolean. You can keep records based on a condition or remove records based on a condition.

If you select to keep rows based on a condition, all records that don’t meet the condition are removed. Only the records that meet the condition remain in the dataset. For example, if you select value is and enter Customer, Wrangler keeps all records with the string Customer and removes all records with other values, such as Lead.

If you select to remove records based on a condition, all records that meet the condition are removed from the dataset. For example, if you select to remove all value is empty, Wrangler removes all records with empty or null values in a column.

You can filter records based on the following conditions:

Condition	Description

Condition	Description
value is empty	Keeps or removes all values with empty values.
value is	Keeps or removes all values with the exact specified value. For columns of type String, you can choose to Ignore case. The default is to include case in the condition.
value contains	Keeps or removes all values that contain the specified value.
value starts with	Keeps or removes all values that start with the specified value.
value ends with	Keeps or removes all values that end with the specified value.
matches regex	Keeps or removes all values that match the regular expression.
custom condition	Keeps or removes all values that match the custom condition.

To keep or remove records based on a condition, follow these steps:

Click the drop-down button next to the column name.
Click Filter and then select the transformation you want to perform.
The values change based on the filter. Wrangler adds the filter-rows-on directive to the recipe. When you run the data pipeline, the transformation is applied to all values in the column.