Pow¶

Overview¶

The Pow expression computes the value of the first argument raised to the power of the second argument. It performs exponentiation using Java's StrictMath.pow function to ensure consistent results across platforms.

Syntax¶

POWER(base, exponent)
-- or
POW(base, exponent)

pow(base, exponent)

Arguments¶

Argument	Type	Description
base	Numeric	The base value to be raised to a power
exponent	Numeric	The exponent to which the base is raised

Return Type¶

Double - returns a double-precision floating-point value representing the result of base^exponent.

Supported Data Types¶

Numeric types (byte, short, int, long, float, double, decimal)
Inputs are converted to double for computation

Algorithm¶

Converts both input arguments to double-precision floating-point values
Delegates computation to Java's StrictMath.pow(double a, double b) method
StrictMath.pow ensures IEEE 754 compliance and consistent results across platforms
Returns the computed result as a double value
Inherits null handling from parent BinaryMathExpression class

Partitioning Behavior¶

This expression preserves partitioning behavior:

Does not require data shuffling between partitions
Can be computed independently on each partition
Does not affect data distribution or partitioning scheme

Edge Cases¶

Returns null if either input argument is null
Returns 1.0 when exponent is 0 (including 0^0 = 1.0)
Returns positive infinity for overflow conditions
Returns 0.0 for underflow conditions
Follows IEEE 754 standards for special cases (NaN, infinity handling)
Negative base with fractional exponent may return NaN

Code Generation¶

This expression supports Tungsten code generation:

Implements doGenCode method for optimized code generation
Generates direct calls to java.lang.StrictMath.pow in generated code
Avoids interpreted execution overhead in code-generated queries

Examples¶

-- Basic exponentiation
SELECT POWER(2, 3);
-- Returns: 8.0

-- Square root using fractional exponent
SELECT POWER(16, 0.5);
-- Returns: 4.0

-- Negative exponent
SELECT POWER(2, -3);
-- Returns: 0.125

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

df.select(pow(col("base_column"), col("exponent_column")))

// With literal values
df.select(pow(col("value"), lit(2)))