Skip to main content

A companion library that enhance your functional superpowers.

Project description

Sidekick is a library that gives you functional superpowers.

Sidekick implements a functional standard library of functions and types designed to make functional programming more pleasant in Python. It uses modern Python features and requires at least Python 3.6+.

Sidekick emphasizes working with immutable types and iterators. It also introduces a special embedded syntax for handling function composition and easy creation of new functions.

Quick start

Install it from pip…

$ pip install sidekick

… and import a few important names

>>> from sidekick.all import sk, op, X, Y, L, N

The examples bellow show a taste of what you can do.

Fibonacci numbers

Let us start with the classics. This produces an infinite sequences of Fibonacci numbers

>>> fibonacci = sk.iterate_past((X + Y), [1, 1])
>>> fibonacci | L[:10]
[1, 1, 2, 3, 5, 8, 13, 21, 34, 55]


iterate_past creates an infinite iterator that generates each number by applying the function in the first argument to the last n elements generated by the sequence. n is given by the size of the initial sequence, which in our example is 2. The Fibonacci sequence is obviously created by always adding the last two elements.

We construct our summation function using the magic objects X and Y. They are function factories that return lambdas corresponding to the expression in which they participate. Hence, X + 1 corresponds to lambda x: x +1 and X + Y to lambda x, y: x + y. If you are familiar to any language in the Haskell family, it should resemble the operator section syntax <>.

In a similar vein, the L magic object that wraps many useful operations on lists. In the example above, L[:10] creates a function that return a list slice in the range of 0 to 10 (not included). L[0] would create a function that fetches the first element, L[::2] would fetch every two elements, and so on.

Finally, the pipe notation passes the argument on the left to the function on the right. This only works on sidekick enabled functions and mimics the pipe in the unix shell. Maybe someday Python can have a native pipe operator like other functional languages <>.

Golden ratio

The snippet above only consumes the first 10 Fibonacci numbers. Let us continue to walk the sequence to find a good approximation to the golden ratio.

>>> ratios = (y / x for (x, y) in sk.window(2, fibonacci))
>>> sk.until_convergence((X == Y), ratios) | sk.last


window produces a sliding window of n elements from the original sequence, e.g., sk.window(2, [1, 2, 3, 4]) ==> (1, 2), (2, 3), (3, 4). The generator comprehension then compute ratios using those consecutive elements.

Finally, the second line iterate over a sequence until the predicate function in the first argument of sk.until_convergence returns True. Rather than setting up some small interval of variation, we test for equality and wait for the difference between two evaluations be smaller than floating point precision.

Euler number

The following snippet uses Taylor formula for exponentials $exp(x) = sum_{n=0}^{infty} frac {x^n} {n!}$to compute the Euler number

>>> factorials = sk.iterate_indexed((X * Y), 1, start=1)
>>> partial_sums = sk.sums(map((1 / X), factorials))
>>> sk.until_convergence((X == Y), partial_sums) | sk.last


iterate_indexed iterates a function f(i, x) passing both the index of iteration and the last evaluation of x to generate the next result. By writing write down a few examples it is easy to see that the given arguments produce a sequence of factorials.

The second step creates a sequence of partial sums of the reciprocal of each factorial. Finally, we iterate until convergence, testing if two consecutive elements are equal.

Sieve of Eratosthenes

The Sieve of Eratosthenes <> is a simple algorithm for selecting all primes in an list of consecutive integers. The list must start with the first prime p (a.k.a., 2), and proceed by excluding every p element. The next valid number will be a prime p’. The procedure is repeated with each new prime until reaching the end of the list.

We will do it like so, except that the initial list of numbers is infinite.

>>> def sieve(nums):
...     p, nums = sk.uncons(nums)
...     yield p
...     yield from sieve(n for n in nums if n % p != 0)
>>> primes = sieve(N[2, 3, ...])
>>> primes | L[:10]
[2, 3, 5, 7, 11, 13, 17, 19, 23, 29]


The fist line in the sieve function uses uncons to extract the first element and return an iterator over the remaining ones. As we described before, the first element is a prime, so we just yield it. The last line of the function applies the sieve to a sequence that excludes every multiple of p.

Finally, we call sieve with the numbers 2, 3, .... The numbers are created by the special object N, specialized in creating numeric sequences. It is very flexible, and in the example above it creates natural numbers starting from 2 and proceed indefinitely in steps of 1. In fact, we could easily make our code operate twice as fast simply by initializing the sieve with N[2, 3, 5, ...] so it would moves in steps of two rather than one. This would avoid checking even numbers which we known in advance not be primes.

See also

Sidekick is heavily inspired by other libraries and functional programming languages. Most notably,

  • toolz: excellent utility library focused on handling iterators.
  • placeholder,, funcy, Pyrsistent: other functional programming libraries for Python.
  • Haskell: an essential inspiration to functional programming. You will see many ideas stolen directly from Haskell. If you want to learn Haskell, however, I recommend learning Elm first ;)
  • Clojure and Elixir: inspiration for many parts of the API.
  • Lodash: a practical functional Javascript library.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for sidekick, version 0.7.0
Filename, size File type Python version Upload date Hashes
Filename, size sidekick-0.7.0-py3-none-any.whl (111.9 kB) File type Wheel Python version py3 Upload date Hashes View
Filename, size sidekick-0.7.0.tar.gz (143.9 kB) File type Source Python version None Upload date Hashes View

Supported by

Pingdom Pingdom Monitoring Google Google Object Storage and Download Analytics Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page