Dataset Access
Dataset Admin Process Flow
The following graphic outlines the sequence of steps required for dataset administration. As an admin, you create groups, add users to those groups, and then assign permissions.
Dataset Access
Auto Insights uses Groups to provide access to a dataset. For each dataset, you can set the following access for Viewers and Owners in that group.
Capability | Viewer | Owner |
|---|---|---|
View Dataset and Insights | ✓ | ✓ |
Configure Dataset Settings | ✓ |
To set group dataset access, go to Admin Portal > Groups. Select the group to edit, then select Viewer & Owner or Viewer from the dropdown. See the next section to learn about Row-based permissions.
Dataset Access for Synthetic Data from Playbooks
Synthetic datasets can be created by non-admin users as part of the Playbooks feature making them the owner of the dataset. Because non-admin users do not have access to the Admin portal and therefore can not grant access to their datasets to other users.
In this case, if a non-admin creates a Mission based on a synthetic dataset as part of the Playbook process and then wants to share the Mission and the dataset with someone else, the request for the dataset access goes to the Organisation Admin who can grant access to all datasets within the organization.
Row-Based Permissions
Set row-based permissions to limit access to underlying data. This is helpful when you have data across many teams. You can set permissions for users, so they can view only data within their respective teams. For example, you could limit a sales manager's view to only their region.
To add row-based permissions...
Go to Admin Portal > Groups.
Select the group you want to assign permission.
Find Row-based permissions under Assigned datasets.
Select the plus sign icon.
Select a column and value from the dropdown menus. You can only select segment fields.
Select Add.
Note
You can not set row-based permissions for company-wide or sample dataset types. If a private dataset with row-based permissions is changed to the company-wide or sample type, Auto Insightsignores the row-based permissions.
Example
Here's an example of row-based permissions where field = Country and value = Australia. This means that only the colored rows are considered when viewing insights for the users that have this row-based permission:
How are multiple row-based permissions handled?
Row-based permissions have very similar functionality to filters (available in the query bar).
When there are multiple row-based permissions for one field, Auto Insights treats them as a logical OR operation.
Example: {field=Team, value=Auto Insights}, {field=Team, value=Designer} means that the user is able to see rows for Auto Insights or Designer.
When there are multiple row-based permissions for different fields, Auto Insights treats them as a logical AND operation.
Example: {field=Team, value=Auto Insights}, {field=Country, value=United States} means that the user only sees rows where Country is United States AND Team is Auto Insights , which yields no rows in the given table example.
How are row-based permissions for one dataset handled across multiple groups?
Auto Insights combines row-based permissions across multiple groups for 1 dataset. Then, it follows the same behavior as multiple row-based permissions.
If a user belongs to a group with no row-based permissions and a group with row-based permissions for a particular dataset, the more restrictive group applies.
Example:
Group 1
Dataset 1
{field=Team, value=Auto Insights}
Group 2
Dataset 1
No row-based permissions
In this case, Auto Insights applies the row-based permissions from Group 1.
Show Underlying Data
This feature allows users to see the underlying line-level raw data for a specific insight shown in Auto Insights to get all the information about this particular insight. This feature is available in Missions and Search.
Note
This feature has to be turned on by an Org Admin.
Turn On the Underlying Data Feature
To turn on this feature for your company...
Go to your avatar in the top right corner, and select Admin Portal. Make sure you are an Org Admin. When you’re logged into the Admin Portal, your role type is displayed next to your name in the top right corner.
Select Feature Access.
Turn on the toggle next to Underlying Data.
Now, dataset owners can allow users to preview and download the underlying raw data for a specific insight on a per dataset level.
The dataset owner can also disable this feature for a dataset. Then users can’t preview or download the underlying data. Users can't download the entire dataset even if preview and download are enabled. It is only a subset of data tied to a specific insight.
How-To for Dataset Owners
On the Datasets page, select the 3-dot menu next to your dataset, and select Configure dataset.
Navigate to Advanced settings in the left navigation bar.
Go to the Underlying data section, and select Enabled from the dropdown.
Now, users can preview and download the underlying raw data for a specific insight on a per dataset level.
On the Datasets page, select the 3-dot menu next to your dataset, and select Configure dataset.
Navigate to Advanced settings in the left navigation bar.
Go to the Underlying data settings section, and select Disabled from the dropdown.
Once disabled, users can’t preview or download the underlying data. Users can't download the entire dataset even if preview and download are enabled. It is only a subset of data tied to a specific insight.
How-To for Users
Go to a Mission page.
There are 2 ways to see underlying data:
Select the ellipsis icon (...) in a table row, then select Show underlying data.
Select the show underlying data icon.
A preview window opens with the line-level table view of the raw data (max 1,000 rows).
(Optional) Download the data in CSV format (max 100,000 rows).
Go to Search.
There are 2 ways to see underlying data:
Select the ellipsis icon (...) in a table row, then select Show underlying data.
Select the show underlying data icon.
A preview window opens with the line-level table view of the raw data (max 1,000 rows).
(Optional) Download the data in CSV format (max 100,000 rows).
FAQ for Users
What are the limitations?
Preview: 1,000 rows
Download 1 CSV: 100,000 rows and 50 columns
Download 2 CSVs: 50,000 rows and 50 columns each
What about when a user has a custom measure, count, or average selected?
AAI still shows the underlying data but will display a warning message that those measures are not 1:1 displayed in the raw data as they require additional calculation on top of the raw data. AAI performs those calculations upon ingestion of data.
What happens when the data comes from 2 different datasets (for example, Common Causes)?
Upon download, you will receive 2 CSV files, each containing a maximum of 50,000 rows. In the preview, you will see two tables.
Does this work for goals (target) datasets?
No. You can't preview or download the underlying data for goals (target) datasets.
Is a disabled column visible for preview or download?
If the dataset owner disabled a column during ingestion, it is not visible for preview or download.