minValue uses Double.MIN_VALUE instead of Double.NEGATIVE_INFINITY

[carloea2]

What happened?

In AggregationOperation.scala, the helper minValue uses Double.MIN_VALUE for AttributeType.DOUBLE. In Java/Scala, Double.MIN_VALUE is the smallest positive non-zero double, not the most-negative value. This makes the “minimum seed” for DOUBLE aggregations incorrect when the intended behavior is to represent the lowest possible value (per discussion #4049, the intent is negative infinity).

texera/common/workflow-operator/src/main/scala/org/apache/amber/operator/aggregate/AggregationOperation.scala

Lines 335 to 345 in 07c35d0

	private def minValue(attributeType: AttributeType): Object =
	attributeType match {
	case AttributeType.INTEGER => Integer.MIN_VALUE.asInstanceOf[Object]
	case AttributeType.DOUBLE => java.lang.Double.MIN_VALUE.asInstanceOf[Object]
	case AttributeType.LONG => java.lang.Long.MIN_VALUE.asInstanceOf[Object]
	case AttributeType.TIMESTAMP => new Timestamp(0)
	case _ =>
	throw new UnsupportedOperationException(
	"Unsupported attribute type for min value: " + attributeType
	)
	}

How to reproduce?

1. Open texera/common/workflow-operator/src/main/scala/org/apache/amber/operator/aggregate/AggregationOperation.scala.
2. Locate minValue(attributeType: AttributeType) (around the case AttributeType.DOUBLE branch).
3. Note it returns java.lang.Double.MIN_VALUE.
4. Expected: the value should be Double.NEGATIVE_INFINITY.

Related discussion: #4049

Version

1.1.0-incubating (Pre-release/Master)

Commit Hash (Optional)

No response

What browsers are you seeing the problem on?

No response

Relevant log output