Multikey Multivalue Non Deterministic python dictionary

Question

There is already a multi key dict in python and also a multivalued dict. I needed a python dictionary which is both:

example:

# probabilistically fetch a         


        

Answer 1

the single output value should be probabilistically determined (fuzzy) based on a rule from keys eg:in above case rule could be if keys have both "red" and "blue" then return "baloon" 80% of time if only blue then return "toy" 15% of time else "car" 5% of time.

Bare in mind your case analysis is not complete, and it's ambiguous, but you can do the following "in spirit" (fleshing out the desired results):

import random

def randomly_return(*colors):
    colors = set(*colors)
    if 'red' in colors and 'blue' in colors:
        if random.random() < 0.8:  # 80 % of the time
            return "baloon"

    if 'blue' in colors and len(colors) == 1:  # only blue in colors
        if random.random() < 0.15:
            return "toy"
        else:
            if random.random() < 0.05:
                return "car"

# other cases to consider

---

I would keep this as a function, because it is a function! But if you insist to make it dict-like, then python let's you do this by overriding __getitem__ (IMO it's not pythonic).

class RandomlyReturn(object):
    def __getitem__(self, *colors):
        return randomly_return(*colors)

>>> r = RandomlyReturn()
>>> r["red", "blue"]  # 80% of the time it'll return "baloon"
"baloon"

---

From your clarification, OP wants to pass and generate:

randreturn((haseither(red,blue),baloon:0.8),((hasonly(blue),toy:0.15)),(default(‌​),car:0.05)))

you want to generate a function as follows:

funcs = {"haseither": lambda needles, haystack: any(n in haystack for n in needles),
         "hasonly": lambda needles, haystack: len(needles) == 1 and needles[1] in haystack}

def make_random_return(crits, default):
    def random_return(*colors):
        colors = set(*colors)
        for c in crits:
            if funcs[c["func"]](c["args"], colors) and random.random() > c["with_prob"]:
                return c["return_value"]
        return default
    return random_return

where the crit and default in this case would be:

crit = [{"func": "haseither", "args": ("red", "blue"), "return_value": "baloon", "with_prob": 0.8}, ...]
default = "car"  # ??
my_random_return = make_random_return(crits, default)

As I say, your probabilities are ambiguous/don't add up, so you're most likely going to need to tweak this...

You can extend the class definition by passing crit and default upon instantiation:

class RandomlyReturn(object):
    def __init__(self, crit, default):
        self.randomly_return = make_random_return(crit, default)
    def __getitem__(self, *colors):
        return self.randomly_return(*colors)

>>> r = RandomlyReturn(crit, default)
>>> r["red", "blue"]  # 80% of the time it'll return "baloon"
"baloon"