MultiplyDTInterval¶

Overview¶

The MultiplyDTInterval expression multiplies a day-time interval by a numeric value, scaling the duration proportionally. This is a binary expression that performs null-safe multiplication between interval and numeric types with proper overflow handling and rounding.

Syntax¶

-- SQL syntax
day_time_interval * numeric_value

-- Function form
INTERVAL '1 2:3:4.567' DAY TO SECOND * 2.5

Arguments¶

Argument	Type	Description
`interval`	DayTimeIntervalType	The day-time interval to be multiplied (left operand)
`num`	NumericType	The numeric multiplier value (right operand)

Return Type¶

Returns DayTimeIntervalType() - a day-time interval representing the scaled duration.

Supported Data Types¶

Left operand: DayTimeIntervalType only
Right operand: All numeric types including:
Integral types (Byte, Short, Int, Long)
Fractional types (Float, Double)
Decimal types with arbitrary precision

Algorithm¶

Intervals are internally stored as microsecond values (Long)
For integral types: Uses Math.multiplyExact() to detect overflow
For decimal types: Converts to Decimal, multiplies, then rounds using HALF_UP mode
For fractional types: Uses DoubleMath.roundToLong() with HALF_UP rounding
All operations preserve microsecond precision in the final result

Partitioning Behavior¶

Preserves partitioning: Yes, this is a row-level transformation
Requires shuffle: No, operates independently on each row
Partition-wise operation: Can be executed within each partition without data movement

Edge Cases¶

Null handling: Null-intolerant expression - returns null if either operand is null
Overflow behavior:
Integral multiplication throws ArithmeticException on overflow via Math.multiplyExact()
Decimal operations may throw ArithmeticException on scale conversion
Fractional operations may produce Infinity or lose precision
Rounding: Non-integral results are rounded to nearest microsecond using HALF_UP mode
Zero multiplication: Results in zero-duration interval
Negative multiplication: Produces negative intervals (reverse direction)

Code Generation¶

Supports full Tungsten code generation with type-specific optimizations: - Integral types: Direct java.lang.Math.multiplyExact() call - Decimal types: Inline Decimal arithmetic with precision handling
- Fractional types: Optimized double arithmetic with DoubleMath.roundToLong()

Examples¶

-- Basic interval multiplication
SELECT INTERVAL '2 10:30:45' DAY TO SECOND * 3;
-- Result: INTERVAL '7 07:32:15' DAY TO SECOND

-- Fractional multiplication with rounding
SELECT INTERVAL '1 12:00:00' DAY TO SECOND * 1.5;
-- Result: INTERVAL '2 06:00:00' DAY TO SECOND

-- Negative multiplication
SELECT INTERVAL '5 08:30:00' DAY TO SECOND * -0.5;
-- Result: INTERVAL '-2 -12:-15:00' DAY TO SECOND

// DataFrame API usage
import org.apache.spark.sql.functions._

df.select(col("day_interval") * lit(2.5))

// With explicit casting if needed
df.select(col("day_interval") * col("multiplier").cast("double"))