You can create calculated fields that mask the data in some way.
But I have hundreds of employee name.....any other way..?
Well I would do calculated something like this:
Left ([Full Name, 2) + Right([Home Address],3) + Left([Spouse Name}, 3)
Result example: Sactewen
Something that will create some kind of masked identifier where only you know the logic and they couldn't figure it out. This is what I do if I need to do something quickly. Then just replace the calculated field with the name field in your workbook.
Create a CF like INT([employee name]) might work. Never done it in Tableau though. I'd recommend pre-processing. I used to have a VBA script that would auto redact reports, I'll see if I can dig it out.
Jonathan Drummey has some good suggestions here, as well:
While most scenarios can be reproduced in the sample data provided by Tableau, for new users, I think this can be time consuming and confusing. If there was an easy data to scramble some fields, or a secure way to share with specified members in the community, I feel I could get more help in addition to getting more work done because I'm not focused on re-building a sample workbook for too long.
These are just my thoughts, as a relatively new user of Tableau. The forums have been the most educational place for me, in terms of learning Tableau--and I understand why sample workbooks are helpful in solving problems here. I just think it could help save time if there was a more efficient way to dummy data.
I also understand this is a complex problem and admittedly, as I learn more, the process of dummying the data becomes more easily accomplished.
Some excellent examples and thoughts from everyone. I just wanted to add that if you are using a calculated field or blending -- you'll be fine as long as you are just showing the dashboards. But if you are allowing anyone to actually have the workbook or view it on a Tableau Server where they can download it or export the data, they'll be able to get to the underlying data and that will include any real data, even if it was masked for show.
Of course, if you've used a join to get fake data columns and then hidden the real fields prior to an extract, you'll have eliminated that issue.
1. Extract the data to a .csv file or Excel file.
2. Use Talend ETL job to do the data disguisement
I've been looking for this! An ideal solution would be to have an option under formatting of dimensions or measures to simply number the items within a dimension.
My example is that I report clinical data by physicians and don't always want to name names - especially at conferences. The ideal would be a formatting option which automatically replaced names from the Physician dimension with, for example, "Physician 1, Physician 2..." allowing them to appear as separate marks and making it clear that the viewer was seeing a comparison across physicians, and keep a colour legend.