Max Number of Fields in Predictive Grouping?

Question

Is there a maximum amount of fields that can be selected to cluster using the predictive grouping tools?

I am trying to use the predictive clustering tools to find 4-6 clusters amongst 577 fields and it is giving me this error -Error: K-Centroids Diagnostics (7): K-Centroids Diagnostics: Error: string at line 1 containing Unicode escapes not in this locale and also ErrorLink: Tool #15: Tool #6: No valid fields were selected. The values in each field are rankings from 0-200 and are doubles.

Any help would be appreciated! Thanks!

CharlieS · Answer

If all your field values are doubles, then this seems it could be a field name issue rather than a limitation on the number of fields. Try using a RecordID and a Dynamic Rename to rename all your fields with some integers to test and see if a field name is the problem.

Another topic with mentioning is that using 577 fields might not be the best course of action. I would consider performing a principle component analysis to reduce dimentionality before the cluster assignment process. Here's a note from the documentation on the Principal Components Tool:

"Principal components can be used instead of the original fields in predictive models, avoiding the problems that can occur when highly correlated variables are used, but at the cost of making model interpretation more difficult. In addition, the method can be used to determine which groups of fields are likely to be jointly highly related to one another, and help guide decisions in which fields to omit from a predictive model. Finally, the ability to "collapse" a large number of fields into a small number of principal components is often a benefit in visualizing relationships in the data."

https://help.alteryx.com/20213/designer/principal-components-tool

mceleavey · Answer

Hi @SBuckingham ,

We'll need to see your data and the workflow. Strip out any identifying data columns and just provide what's being used in the tool.

M.