Clustering of 1.5 lakhs external records

[soumyadip28]

I am facing difficulties in clustering data. I have some 1.5 lakhs records which I have uploaded by creating custom table and transform map to target table by matching the data types in each table. When the count of the records were less approx. 5000 , the clustering was taking place but when the count got increased the  clustering is failing saying "NSE0016:Failed to execute component clusteringValidator : NCE0077:No text column in Input features: ['u_current_service']". Please help me fix this issue.

[Naval Singh]

As the number of rows increased, there might be a row in the field which might be null or numerical etc. Clean the column. 

[Brian Bakker]

@soumyadip28 

If u_current_service is a Reference field, then in the Clustering solution definition, you will need to dot-walk to the 'name' field for this Input field on this Clustering solution, so that it looks like 'u_current_service.name', to ensure you train the solution on the text, and not the sys_id.