HyperUnique columns in Pivot

Hey Guys, I’m sure this is a dumb question, but what is the expression for a hyperunique measure in the config.yaml. Tried to dig through plywood but didn’t see anything that stood out. Thanks! - Lindsay

Hi Lindsay,

The expression is: “$main.countDistinct($your_metric_name)”.

Here is an example: https://github.com/implydata/pivot/blob/master/config.yaml.sample#L161

Please reply here if that does not work, there might be one more tweak that you have to do.

Hey Vadim, I tried that—it just returned 1 for each row. We’re using hyperUnique in the ingestion spec.

Saw your video on Plywood today btw. Great stuff.

Thank you!
I am sorry it is not working for you, can you please answer the following questions for me and we will get this resolved ASAP.

What is the name of the metric?

What version of Druid are you running: 0.8.1? earlier? or IAP Druid?

What version of Pivot are you using? (run pivot --version) or are you running it through bard?

If you do not want to share these details publicly you can also email me the answers as well.

I believe I know where the problem is and we should get this resolved very soon.

Even I faced similar problem. Pivot shows 1.0 for hyperUnique metrics. I tried this in IAP and also on druid 0.8.1. Anyway, you can define hyperUnique metric in customAggregations ( eg. here- https://github.com/implydata/pivot/blob/master/config.yaml.sample#L161) and it shows the correct value.

I have published a new version of Pivot just now (0.5.10).
I believe I have fixed the problem and the same config you have now should work as expected in the new version of Pivot.

Of course if it does not work the custom customAggregations that Saksham suggested would definitely work and are a food fallback.