Skip to content

squirrelinhell/mandalka

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

35 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Install

pip install --upgrade git+https://github.com/squirrelinhell/mandalka.git

What is it?

Mandalka makes Python objects unique, based on arguments passed to __init__. So this kind of code will run fast:

import mandalka

@mandalka.node
class Fib:
    def __init__(self, n):
        if n <= 1:
            self.n = n
        else:
            self.n = Fib(n-1).n + Fib(n-2).n

print(Fib(100).n) # prints "354224848179261915075"

It also provides lazy evaluation, which makes it possible to transparently cache and re-use data:

import os
import numpy as np

@mandalka.node
class Data:
    def __init__(self, size):
        # This code will not run if Model doesn't need it
        print("Computing...")
        self.squares = np.square(np.arange(size))

@mandalka.node
class Model:
    def __init__(self, data):
        path = mandalka.unique_id(self) + ".npy"
        if os.path.exists(path):
            self.weights = np.load(path)
        else:
            self.weights = data.squares * 10
            np.save(path, self.weights)

model = Model(Data(5))

# Prints "Computing..." only for the first time this script is run:
print(model.weights)

Arguments passed to constructors of nodes can only be made out of other nodes and basic Python types (int, str, list, tuple, dict etc.).

The uniqueness of mandalka nodes makes it possible to write code more similar to what one would expect from functional programming, or mathematical notation:

@mandalka.node
class S:
    def __init__(*_):
        pass

two = S(S(0))
three = S(S(S(0)))
assert three == S(two)

def add(a, b):
    counter = 0
    while counter != b:
        a = S(a)
        counter = S(counter)
    return a

five = add(two, three)
print(mandalka.describe(five, -1)) # prints "S(S(S(S(S(0)))))"

License

This code is distributed under the MIT license, as quoted below:

Copyright (c) 2017 SquirrelInHell

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

About

Computational graphs from Python classes

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published