Universal functions in in numpy

By Martin McBride, 2021-01-27
Tags: arrays ufunc universal function vectorisation out
Categories: numpy


The section on vectorisation looked to apply arithmetic operators across a whole array in a single expression. Universal functions allow us to apply mathematical functions across a whole array in a similar way. Universal functions, or ufuncs for short, special NumPy versions of standard maths functions.

Example universal function - sqrt

The sqrt ufunc calculates the square root of each element in an array. For example:

a = np.array([1, 2, 3, 4])
b = np.sqrt(a)

This calculates the square root of each element 1, 2, 3, 4, giving the result:

b = [1.         1.41421356 1.73205081 2.        ]

Of course this can be applied to multi-dimensional arrays too, for example a 2 by 3 array:

a = np.array([[10, 20, 33], [40, 50, 60]])
b = np.sqrt(a)

This again calculates the square root of each element and returns another 2 by 3 array:

b = [[3.16227766 4.47213595 5.74456265]
     [6.32455532 7.07106781 7.74596669]]

Example universal function of two arguments - power

Some ufuncs take two arguments, for example the power function:

a = np.array([5, 10, 5, 10])
b = np.array([2, 2, 3, 3])
c = np.power(a, b)

power(x, y) calculates x to the power y. The function is equivalent to x**y.

So power(5, 2) is 5 squared, or 25, and so on:

b = [  25  100  125 1000]

Summary of ufuncs

There are quite a number of ufuncs, and they are all described in the official NumPy documentation. The main groups of functions are:

  • Maths operations
  • Trigonometric functions
  • Bit manipulation
  • Comparison functions
  • Logical functions
  • Float functions

Additional arguments

ufuncs can be called with additional, optional arguments:

out

Normally, a ufunc creates a new NumPy array to hold its result:

a = np.array([1, 2, 3, 4])
b = np.array([2, 4, 6, 8])
c = a + b

The out parameter allows us to specify an existing array for the output. To use this feature, we must use the add ufunc rather than the + operator:

a = np.array([1, 2, 3, 4])
b = np.array([2, 4, 6, 8])
r = np.zeros_like(a)
np.add(a, b, out=r)

This fills the array r with the result of a + b. The shape of the output array must be compatible with the input arrays.

One case where this is useful is if you want to re-use an existing array, for example to add a to b and leave the result in a. This is particularly useful if the arrays are very large. Here is how to do it:

a = np.array([1, 2, 3, 4])
b = np.array([2, 4, 6, 8])
np.add(a, b, out=a)

See also

If you found this article useful, you might be interested in the book NumPy Recipes or other books by the same author.

Join the PythonInformer Newsletter

Sign up using this form to receive an email when new content is added:

Popular tags

2d arrays abstract data type alignment and angle animation arc array arrays bar chart bar style behavioural pattern bezier curve built-in function callable object chain circle classes clipping close closure cmyk colour combinations comparison operator comprehension context context manager conversion count creational pattern data science data types decorator design pattern device space dictionary drawing duck typing efficiency ellipse else encryption enumerate fill filter font font style for loop formula function function composition function plot functools game development generativepy tutorial generator geometry gif global variable gradient greyscale higher order function hsl html image image processing imagesurface immutable object in operator index inner function input installing iter iterable iterator itertools join l system lambda function latex len lerp line line plot line style linear gradient linspace list list comprehension logical operator lru_cache magic method mandelbrot mandelbrot set map marker style matplotlib monad mutability named parameter numeric python numpy object open operator optimisation optional parameter or pandas partial application path pattern permutations pie chart pil pillow polygon pong positional parameter print product programming paradigms programming techniques pure function python standard library radial gradient range recipes rectangle recursion reduce regular polygon repeat rgb rotation roundrect scaling scatter plot scipy sector segment sequence setup shape singleton slice slicing sound spirograph sprite square str stream string stroke structural pattern subpath symmetric encryption template tex text text metrics tinkerbell fractal transform translation transparency triangle truthy value tuple turtle unpacking user space vectorisation webserver website while loop zip zip_longest

Numpy
Navigation
Links