Using Filters in Lumira Discovery 2.1

[former_member548161]

I have been using Lumira Discovery 2.1 to build data quality visualizations. I am trying to create a chart showing the number of e-mails missing either a '.' OR an '@' symbol. The important word here is OR.

I created a filter using 'Does Not Contain'. When I enter only one parameter I get proper results. That is, Does Not Contain '.' returns all records missing the '.' in the e-mail address. Similarly, Does Not Contain '@' returns all records missing the '@' in the e-mail address. However, Does not contain '.' '@' only returns records missing both the period and the at symbol. The filter is clearly using AND to join the two parameters.

My question is, can I create the filter using OR to join the parameters so that it returns all records missing either a period OR an at symbol?

Thanks.

[TammyPowlas]

Mark - depending on your data source, the query with SQL might work with this - see the Help: https://help.sap.com/viewer/Discovery_User_Guide/e78a678580b7101487253286b0e91070.html

[former_member548161]

I am actually using Query with SQL to acquire the data from a DB2 database. If I add the "filter" to my query then I won't have the remainder of the rows from the table. I guess I could create different queries for each metric but I would like to avoid that by using the filters. But if it's not possible...

The longer I work in Discovery the more limitations I see. I haven't investigated Lumira Designer yet. Perhaps its advanced capabilities could help??

[TammyPowlas]

With Designer you would need a universe/semantic layer on top of it. I hope others can chime in with some suggestions

[former_member548161]

Thanks Tammy. I was not aware of that requirement for using Designer. I am trying to figure out if I can create a calculation to do what I want but the available functions in Discovery are limited and the logic I would need is escaping me at the moment. I too hope others chime in with ideas.

[former_member548161]

Well, I am going to answer my own question here. I don't believe you can do this with a filter. At least I could not find a way. Instead, I had to create a custom calculation. I created a new Dimension called BAD_EMAIL using the following formula:

if Contain({EMAIL},".") and Contain({EMAIL}, "@") then "GOOD" else "BAD"

Then when I created my chart I used the new Dimension, BAD_EMAIL, as the filter and selected records that were "BAD". If someone knows a better way I'd love to hear it. Also, if anyone knows of a place to see real example of formulas in Lumira I'd love to know .

Thanks.