Creating an Advanced Analytics DCR
This section will walk you through the process of creating an Advanced Analytics DCR.
Your organisation needs an active license and needs to be analytics enabled to create AA DCRs. If you have questions, please contact your customer success representative or support@decentriq.com.
Step 1: Create an Advanced Analytics DCR
- Click on the New DCR button in the sidebar. Depending on your type of access, you might be presented with other types of data clean rooms. If so, select Advanced Analytics.
- Provide a name for your DCR.
- Choose to start from scratch or import an existing DCR configuration which can be downloaded from existing DCRs.
- A new Advanced Analytic DCR in draft has been created. At this point, only you have access to it. To make it visible to other users, you have to publish the DCR (see below).
![Choose DCR type](/img/aa_dcr-imgs/choose_dcr.png)
![Name DCRs](/img/aa_dcr-imgs/name_aadcr.png)
Step 2: Configure the datasets in the DATA tab
Define the datasets to be provided by the data owners:
Structured data
- Add a new "Table" and give it a name
- For each column configure its name, data type, if its values are hashed, if it allows empty values. Note that the schema can be auto-detected from a file provided by you using Import schema.
- Set if the dataset must be provided for computations to run and unique value constraints (optional)
![add table to DCR](/img/aa_dcr-imgs/adding_table_aadcr.png)
Unstructured data
- Add a new "File"
- Set if the dataset must be provided for computations to run and unique value constraints
![add file to DCR](/img/aa_dcr-imgs/adding_file_aadcr.png)
Step 3: Configure computations in the COMPUTATIONS tab
Add computations that will be executed on the previously defined data. The computations can depend on both, the previously defined data as well as the output of computations themselves. This allows chaining of computations. The computations can only be executed and the results retrieved by participants who have been given the corresponding permission (see below).
![add new computation to DCR](/img/aa_dcr-imgs/new_node_aadcr.png)
See Computations for descriptions of the available computations.
Step 4: Test the DCR configuration
You can upload test data and run the computations over it.
Documentation coming soon.
Step 5: Define permissions in the PERMISSIONS tab
Configure which participants will be invited to the collaboration, and assign them permissions.
Add new participants by typing in their email. They will receive email-based invitations as soon as the DCR is published.
Use the dropdown boxes to assign data owner permissions. If a participant is data owner of a dataset, they can provision data to it. It is possible, that multiple participants are data owner of the same dataset. In this case, either of them can provision data to it.
Use the dropdown boxes to assign analyst permissions. If a participant is analyst of a computation, they can run and retrieve the results of it.
By default, Advanced Analytics DCRs are interactive, allowing participants to request new computations which have to be approved by the data owners. To prevent any modifications after publishing, check the This data clean room is immutable option.
![add file to DCR](/img/aa_dcr-imgs/new_permission_aadcr.png)
Step 6: Publish the DCR
To publish the DCR, click on Publish button in the top-right corner. By default, this test-runs the DCR with the provided test datasets (empty if none provided) to ensure the correctness of the computations. Note that running the computations on empty datasets may result in errors. To avoid this, choose Publish without validation
After publishing, participants are invited to the DCR and can begin collaborating. See the Using an Advanced Analytics DCR as a data analyst and/or the Using an Advanced Analytics DCR as a data owner section for the next steps.
![publish DCRs](/img/aa_dcr-imgs/publishing_aadcr.png)