689741_en-US

liliconcom

Cross-column duplicate run size in Explore Patterns

Hello, I am misunderstanding how to use cross-column duplicate run size in Explore Patterns - can anyone offer suggestions?  I followed the sample data steps from an example in the Help menu, and I don't know what I'm doing differently.

 

MAC addresses correspond to individual sensors.  I'm looking for streaks of the same Event Code at the same Date/Time.

 

There are about 30 duplicates highlighted in the table below.  However, the report says "No duplicates were found across columns. No duplicates found with run length of at least 5"

 

Thanks, Paul

 

Data Exploration and VisualizationRe: Cross-column duplicate run size in Explore Patterns

Duplicates Across Columns is looking for runs of rows where two or more columns have the same value in the same row.

 

Here's an example:

 

The red cells in rows 603 and 604 have the same values in the columns Erythrocytes and Potassium.

 

I don't think Explore Patterns has an option that would detect runs of rows across columns.

 

Instead you could create a new column that concatenates the values from your two columns and then the Univariate Runs would detect them.

 

Alternatively you could use Tables->Summary and group by your Time and EventCode columns and you'll get a table (linked to the original table) that groups time and event codes.

 

Re: Cross-column duplicate run size in Explore Patterns

Thank you, Jeff, for clearing up my misconception.  Your suggestions are both helpful.