Apache Pinot

Sorry, another date time field spec question: can you have multiple date time columns bucketed on different granularities based on one source column?

yes. multiple date time field spec are allowed. can do what you wrote

Oh, so the name corresponds to the "output" name and the transform function will use the source column?

when you write transform functions like,
y = f(x)
z = f(x)
`x` is expected to be in the source data. `y` and `z` are expected to be in the Pinot schema. Your final Pinot data will contain y and z

Thanks! Would this work: `floor(create_time_epoch_seconds/ 60)`?

create_time_epoch_seconds has to be in the source. it cannot be the result of another transform function

And also we get errors when doing something like `60 * toEpochMinutes(fromEpochSeconds(create_time_epoch_seconds)),`

```"message": "QueryExecutionError:\njava.lang.ClassCastException: java.lang.Long cannot be cast to java.lang.Double\n\tat org.apache.pinot.core.operator.transform.function.ScalarTransformFunctionWrapper.transformToDoubleValuesSV(ScalarTransformFunctionWrapper.java:236)\n\tat org.apache.pinot.core.operator.transform.function.MultiplicationTransformFunction.transformToDoubleValuesSV(MultiplicationTransformFunction.java:76)\n\tat```

Is that a bug? If so I can help to fix it

As long as the first one works I think that's sufficient to bucket as we need, was more just curious about the error, we got a few of them.

Might be easier to open a github issue for those or just paste them?

Also don't want to bother you too much - you've been so helpful to us.

<@UDT7GFEG6> is your post aggregation merged? Will this be supported now ^^ ?

<@UPK7Y0X0A> Are you using the latest version of Pinot? Type casting should be supported now

`0.5.0` won't include the fix as the fix is checked in very recently

Do you recommend checking out the latest or waiting for 0.6.0?

You may try out the current master in a testing environment and see whether the issue is fixed

Release `0.6.0` might be cut at around end of the year, so you can decide based on the urgency of the fix