Python math.nextafter() Function
Python math ModuleIn floating-point calculations, understandingadjacent floating-point numbersThe relationship between them is very important. Floating-point numbers in computers are discrete, and there is a minimum interval between two adjacent floating-point numbers.
math.nextafter()is a function introduced in Python 3.9, used to return the value after advancing a specified number of floating-point steps from x toward y.
Word Meaning： nextafterIt means "after the next".
Basic Syntax and Parameters
Syntax Format
import math math.nextafter(x, y[, steps])
Parameter Description
- x: starting floating-point number
- y: target direction
- steps(optional): number of steps, default is 1
Return Value
Return the floating-point number after advancing steps steps from x in the direction of y
Examples
Example 1: Get adjacent floating-point numbers
Example
import math
next_val = math.nextafter(1.0, math.inf)
print(f"The next floating-point number after 1.0: {next_val}")
print(f"Difference: {next_val - 1.0}")
prev_val = math.nextafter(1.0, -math.inf)
print(f"\n"The previous floating-point number before 1.0: {prev_val}")
print(f"Difference: {1.0 - prev_val}")
next_val = math.nextafter(1.0, math.inf)
print(f"The next floating-point number after 1.0: {next_val}")
print(f"Difference: {next_val - 1.0}")
prev_val = math.nextafter(1.0, -math.inf)
print(f"\n"The previous floating-point number before 1.0: {prev_val}")
print(f"Difference: {1.0 - prev_val}")
Output:
1.0 的下一个浮点数: 1.0000000000000002 差值: 2.220446049250313e-16 1.0 的前一个浮点数: 0.9999999999999999 差值: 1.1102230246251565e-16
Example 2: steps parameter
Example
import math
print("Forward 10 steps:", math.nextafter(1.0, math.inf, 10))
print("Negative step count:", math.nextafter(1.0, -math.inf, -5))
print("Forward 10 steps:", math.nextafter(1.0, math.inf, 10))
print("Negative step count:", math.nextafter(1.0, -math.inf, -5))
Output:
前进 10 步: 1.0000000000000007 负数步数: 0.9999999999999996
Example 3: Special Values
Example
import math
print("Infinity:", math.nextafter(math.inf, 0))
print("After 0:", math.nextafter(0.0, 1.0))
print("Negative zero:", math.nextafter(-0.0, 1.0))
print("Infinity:", math.nextafter(math.inf, 0))
print("After 0:", math.nextafter(0.0, 1.0))
print("Negative zero:", math.nextafter(-0.0, 1.0))
Output:
无穷大: 1.7976931348623157e+308 0 之后: 5e-324 负零: 5e-324
Example 4: Calculate ULP
Example
import math
for x in [1.0, 2.0, 100.0]:
ulp = math.nextafter(x, math.inf) - x
print(f"ULP({x}) = {ulp}")
for x in [1.0, 2.0, 100.0]:
ulp = math.nextafter(x, math.inf) - x
print(f"ULP({x}) = {ulp}")
Output:
ULP(1.0) = 2.220446049250313e-16 ULP(2.0) = 4.440892098500626e-16 ULP(100.0) = 2.8421709430404007e-14
Notes
- This function is only available in Python 3.9+
- The steps parameter is also new in Python 3.9
- Can be used to understand floating-point precision limitations
Other Extensions