How to calculate the number of distinct incidents in each jurisdiction?

[scottmkirkland]

I have a dataset with incident numbers and their associated Jurisdiction. It is possible that a incident will be listed in multiple jurisdictions. 

I don't want to dedup(incident_number) globally.

I need to count by jurisdiction, but the dedup or distinct count needs to be within each Jurisdiction. 

Any suggestions?

[bowesmana · Accepted solution]

Use the by clause in stats command, e.g.

| stats count by jurisdiction

OR

| stats dc(incident_number) by jurisdiction