Highlambder - there can be only one... lambda to rule them all

Syntactic sugar for Python lambda expressions. Make your anonymous functions shorter, more expressive, and (let's be honest) just more fun.

Image used under fair use, for illustrative and non-commercial purposes. All rights to the character and image belong to their respective owners.

What is Highlambder?

Highlambder is an experimental library that lets you write more elegant and compact code by replacing Python’s lambda syntax with something shorter: a symbolic placeholder that behaves like a function.

In practice, lambda x: x + 1 can be written simply as λ + 1.

Installation

pip install highlambder

Quick Examples

from highlambder import L as λ

λ (10)  # -> 10

(λ + 5) (10)  # -> 15

(λ * 5) (10)  # -> 50

(3 + λ * 2) (10)  # -> 23

(40 / λ / 5) (2)  # -> 4

(10 * λ[1]) ([1, 2, 3])  # -> 20

(-1 + λ * 5 / λ + 1) (13)  # -> 5

(λ * 2 + λ * 4 + λ) (10)  # -> 70

(λ['A'] + λ['B']) ({'A': 3, 'B': 4})  # -> 7

(λ + λ) (2)  # -> 4

map(λ * 2, range(5)) # -> [0, 2, 4, 6, 8]

("It's a Me, " + λ) ('Mario!') # ->  'It's a Me, Mario!'

pandas and NumPy support (partially supported, work-in-progress)

import pandas as pd
import numpy as np
from highlambder import L as λ

# pandas:
s = pd.Series([1, 2, 3, 4])
assert pd.Series.equals(
    2 * s,
    s.map(2 * λ),
)

df = pd.DataFrame({
    'A': [1, 1, 2, 2],
    'B': [5, 6, 7, 8],
    'C': ['banana', 'apple', 'kiwi', 'orange'],
})

assert pd.DataFrame.equals(
    df.assign(D=lambda d: d.A + 20),
    df.assign(D=λ.A + 20),
)

# String operations
assert pd.DataFrame.equals(
    df.assign(D=lambda d: d['C'].str.len() * 2),
    df.assign(D=λ['C'].str.len * 2),
)

# NumPy:
assert (λ + 2)(np.int64(2)) == 4

assert (λ.max)(np.array([3, 4, 5, 6, 7, 8])) == 8

Limitations (for now)

• Only single-argument functions are supported.
• Calling a function passing λ as an arument usually doesn't work. For example: len(λ).

These limitations may be lifted in future versions.

License

MIT