groupBy 후 collect_list 에 여러개의 column 값을 담기

collect_list(struct(???)) 를 사용한다.

import spark.implicits._


case class BaseMetric(
	id: String,
	log_type: String,
	device_type: String,
	clk_count: Long,
	imp_count: Long
)

case class BaseMetricGroup(
	id: String,
	metrics: Seq[BaseMetric]
)

val metricDs: Dataset[BaseMetric] = ???
metricDs
	.groupBy("id")
	.agg(
		collect_list(
			struct(BaseMetric.fields.map(col): _*)
		).as("metrics")
	)
	.as[BaseMetricGroup]