Create a Calculation View with Differential Privacy in SAP HANA Cloud

Using what you learned in Create an SAP HANA Database Project, you will create another database table and load sample data into it but without step by step instructions. Extrapolate what you learned from these same steps in the earlier tutorial to complete this task.

1. Create a new database table named TGT_SALARIES using the hdbtable artifact type in the data folder. Use the following table definition.

   SQL

   Copy

   column table "TGT_SALARIES" (
       ID INTEGER  unique not null COMMENT 'Employee ID',
       SALARY DOUBLE COMMENT 'Salary',
       START_YEAR INTEGER not null COMMENT 'Starting Year',
       GENDER NVARCHAR(1) COMMENT 'Gender',
       REGION NVARCHAR(4) COMMENT 'Region',
       "T-LEVEL" NVARCHAR(200) COMMENT 'T Level',
       PRIMARY KEY ("ID")
   )
   COMMENT 'Employee Salary'
   

2. Create the configuration for the csv upload. In the data/loads folder, create a file called salaryload.hdbtabledata with the following content

   json

   Copy

   {
     "format_version": 1,
     "imports": [{
       "target_table": "TGT_SALARIES",
       "source_data": {
         "data_type": "CSV",
         "file_name": "salarydata.csv",
         "has_header": true,
         "dialect": "HANA",
         "type_config": {
           "delimiter": ","
         }
       }
       }]
   }		
   

3. Download this csv file – https://github.com/SAPDocuments/Tutorials/blob/master/tutorials/hana-cloud-calculation-view-differential-privacy/salarydata.csv into your computer. Upload it into the loads folder using the Upload Files option.

4. Deploy these new Artifacts

   

1. Create a new folder called models under db/src. Create a new Calculation View via the SAP HANA: Create SAP HANA Database Artifact command pallet entry.

   

2. Call it SALARIES_ANONYMIZED and make it a CUBE. Press Create.

   

3. Select the new calculation view file in the Explorer. This will load it into the graphical Calculation View editor. Click on the Projection node and then click on the white canvas to drop it.

   

4. Use the

   
   on the node to add a table as a data source. Choose TGT_SALARIES
   

1. Double click on the Projection_1 node. This will open the mapping. Double click on TGT_SALARIES to add all of the columns to the output

   

2. Connect the Projection_1 node to the Aggregation node.

   

3. Double-click on the name of the node to move all the fields into the output columns.

   

4. Go into the Semantics node and switch START_YEAR and ID to attribute

   

5. Deploy

1. Create a view (artifact type hdbview) in the src/models folder named V_SALARIES.

   

2. Use the following syntax to create a SQL view that uses the Differential Privacy Anonymization approach for the SALARY column.

   SQL

   Copy

   VIEW V_SALARIES (
     ID,
     SALARY,
     START_YEAR,
     GENDER,
     REGION,
     "T-LEVEL"
   ) AS
   SELECT ID,
     SALARY,
     START_YEAR,
     GENDER,
     REGION,
     "T-LEVEL"
   FROM "SALARIES_ANONYMIZED"
   WITH READ ONLY
   WITH ANONYMIZATION (ALGORITHM 'DIFFERENTIAL_PRIVACY'
     PARAMETERS '{"data_change_strategy": "qualified"}'
     COLUMN ID PARAMETERS '{"is_sequence": true}'
     COLUMN SALARY PARAMETERS '{"is_sensitive":true, "epsilon":0.1, "sensitivity":15000}')
   

   For more information about these parameters check the SAP HANA Cloud Data Anonymization Guide

3. If you receive an error that the feature is not supported, this can be safely ignored.

   

4. Deploy

1. Open the Database Explorer for your project

   

2. Open a SQL Console and issue the following statement

   SQL

   Copy

   refresh view V_SALARIES anonymization;
   

3. Execute that statement in the SQL console

   

4. You can now use the normal Open Data on the view to preview the Raw Data.