Github user cloud-fan commented on a diff in the pull request:
https://github.com/apache/spark/pull/20062#discussion_r158717586
--- Diff:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/plans/logical/statsEstimation/EstimationUtils.scala
---
@@ -89,29 +89,29 @@ object EstimationUtils {
}
/**
- * For simplicity we use Decimal to unify operations for data types
whose min/max values can be
+ * For simplicity we use Double to unify operations for data types whose
min/max values can be
* represented as numbers, e.g. Boolean can be represented as 0 (false)
or 1 (true).
* The two methods below are the contract of conversion.
*/
- def toDecimal(value: Any, dataType: DataType): Decimal = {
+ def toDouble(value: Any, dataType: DataType): Double = {
dataType match {
- case _: NumericType | DateType | TimestampType =>
Decimal(value.toString)
- case BooleanType => if (value.asInstanceOf[Boolean]) Decimal(1) else
Decimal(0)
+ case _: NumericType | DateType | TimestampType =>
value.toString.toDouble
+ case BooleanType => if (value.asInstanceOf[Boolean]) 1 else 0
}
}
- def fromDecimal(dec: Decimal, dataType: DataType): Any = {
+ def fromDouble(double: Double, dataType: DataType): Any = {
dataType match {
- case BooleanType => dec.toLong == 1
- case DateType => dec.toInt
- case TimestampType => dec.toLong
- case ByteType => dec.toByte
- case ShortType => dec.toShort
- case IntegerType => dec.toInt
- case LongType => dec.toLong
- case FloatType => dec.toFloat
- case DoubleType => dec.toDouble
- case _: DecimalType => dec
+ case BooleanType => double.toLong == 1
--- End diff --
maybe `toInt`?
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
