Anonymise a dataset

Tip

You can get random company names from:

• https://randommer.io/random-business-names

You can get random person names from:

• http://listofrandomnames.com/index.cfm?generated
• https://www.randomlists.com/random-names?qty=70
• https://www.behindthename.com/random/

1. Export all Dim and Fact tables to Excel from pbix

   1. This will allow us to create a pbix that is completely independent of the original data source
   2. usually, copy table from data view in Power BI works better than from PowerQuery, because not truncated
   3. To copy table, just select it and use context menu on table
   4. Filters etc. work when exporting, in case you already want to limit what's to be anonymised

2. delete calculated columns in Excel

3. anonymise in Excel (see Random (Company) Names)

4. Create a new pbix file for the anonymised data

5. Import all excels into power query

   1. If you filtered some Dim records, you might want to/have to filter the fact records given the dimensions

6. Do transformations

   1. Header names
   2. data types → they must match exactly the original tables

7. save without applying (the next step is not repeatable or undoable)

8. Apply (recheck relationships and visuals)

9. Create Relationships