random.choice from set? python
Question:
I’m working on an AI portion of a guessing game. I want the AI to select a random letter from this list. I’m doing it as a set so I can easily remove letters from the list as they are guessed in the game and are therefore no longer available to be guessed again.
it says set object isn’t indexable. How can I work around this?
import random
aiTurn=True
while aiTurn == True:
allLetters = set(list('abcdefghijklmnopqrstuvwxyz'))
aiGuess=random.choice(allLetters)
print (aiGuess)
Answers:
Note (Oct. 2020): as of v3.9, Python has officially deprecated random.sample() working on sets, with the official guidance being to explicitly convert the set to a list or tuple before passing it in, though this doesn’t solve the efficiency problems.
>>> random.sample(set('abcdefghijklmnopqrstuvwxyz'), 1)
['f']
Documentation: https://docs.python.org/3/library/random.html#random.sample
Note that choosing random elements from a set is extremely inefficient no matter how you do it – it takes time proportional to the size of the set, or worse if the set’s underlying hash table is sparse due to removed elements.
Instead, you should probably use a different data structure that supports this operation efficiently.
You should use random.choice(tuple(myset)), because it’s faster and arguably cleaner looking than random.sample. I wrote the following to test:
import random
import timeit
bigset = set(random.uniform(0,10000) for x in range(10000))
def choose():
random.choice(tuple(bigset))
def sample():
random.sample(bigset,1)[0]
print("random.choice:", timeit.timeit(choose, setup="global bigset", number=10000)) # 1.1082136780023575
print("random.sample:", timeit.timeit(sample, setup="global bigset", number=10000)) # 1.1889629259821959
From the numbers it seems that random.sample takes 7% longer.
You could work around this by using a list instead of a set. You will still be able to remove letters "easily" from the list. Try this, for example:
allLetters = list('abcdefghijklmnopqrstuvwxyz')
aiGuess = random.choice(allLetters)
allLetters.remove(aiGuess)
Another option is to randomly choose the index instead of the letter, which might be slightly faster since we don’t need to search for the element to delete (but I’d question if speed actually matters here?):
allLetters = list('abcdefghijklmnopqrstuvwxyz')
index = random.randint(0, len(allLetters)-1) # Top is inclusive, unlike slices
aiGuess = allLetters[index]
del allLetters[index]
Since the choice list is not very long, you can use random.shuffle the list first. Then iterate each element from the list. This avoid removing element from a list one by one and make your code cleaner.
If you want to get a random element from the set.
a = set()
for i in range(10):
a.add(i)
a.pop() // gives a random element from a set
You can combine a doubly-linked list and a dictionary, to create a set with O(1) random choice.
You could try this:
def random_set_ele(set_: set):
copy = set_
return copy.pop()
test_set = set(["".join(["elem-", str(l)]) for l in range(0, 1000000)])
start = perf_counter()
print(random_set_ele(test_set))
print(perf_counter()-start)
Result:
elem-57221
0.00016391400276916102
The .pop() method for a set randomly extracts and returns an element and pops it out of the set. We make a copy of the set before popping the element so that the original list is not modified.
I’m working on an AI portion of a guessing game. I want the AI to select a random letter from this list. I’m doing it as a set so I can easily remove letters from the list as they are guessed in the game and are therefore no longer available to be guessed again.
it says set object isn’t indexable. How can I work around this?
import random
aiTurn=True
while aiTurn == True:
allLetters = set(list('abcdefghijklmnopqrstuvwxyz'))
aiGuess=random.choice(allLetters)
print (aiGuess)
Note (Oct. 2020): as of v3.9, Python has officially deprecated random.sample() working on sets, with the official guidance being to explicitly convert the set to a list or tuple before passing it in, though this doesn’t solve the efficiency problems.
>>> random.sample(set('abcdefghijklmnopqrstuvwxyz'), 1)
['f']
Documentation: https://docs.python.org/3/library/random.html#random.sample
Note that choosing random elements from a set is extremely inefficient no matter how you do it – it takes time proportional to the size of the set, or worse if the set’s underlying hash table is sparse due to removed elements.
Instead, you should probably use a different data structure that supports this operation efficiently.
You should use random.choice(tuple(myset)), because it’s faster and arguably cleaner looking than random.sample. I wrote the following to test:
import random
import timeit
bigset = set(random.uniform(0,10000) for x in range(10000))
def choose():
random.choice(tuple(bigset))
def sample():
random.sample(bigset,1)[0]
print("random.choice:", timeit.timeit(choose, setup="global bigset", number=10000)) # 1.1082136780023575
print("random.sample:", timeit.timeit(sample, setup="global bigset", number=10000)) # 1.1889629259821959
From the numbers it seems that random.sample takes 7% longer.
You could work around this by using a list instead of a set. You will still be able to remove letters "easily" from the list. Try this, for example:
allLetters = list('abcdefghijklmnopqrstuvwxyz')
aiGuess = random.choice(allLetters)
allLetters.remove(aiGuess)
Another option is to randomly choose the index instead of the letter, which might be slightly faster since we don’t need to search for the element to delete (but I’d question if speed actually matters here?):
allLetters = list('abcdefghijklmnopqrstuvwxyz')
index = random.randint(0, len(allLetters)-1) # Top is inclusive, unlike slices
aiGuess = allLetters[index]
del allLetters[index]
Since the choice list is not very long, you can use random.shuffle the list first. Then iterate each element from the list. This avoid removing element from a list one by one and make your code cleaner.
If you want to get a random element from the set.
a = set()
for i in range(10):
a.add(i)
a.pop() // gives a random element from a set
You can combine a doubly-linked list and a dictionary, to create a set with O(1) random choice.
You could try this:
def random_set_ele(set_: set):
copy = set_
return copy.pop()
test_set = set(["".join(["elem-", str(l)]) for l in range(0, 1000000)])
start = perf_counter()
print(random_set_ele(test_set))
print(perf_counter()-start)
Result:
elem-57221
0.00016391400276916102
The .pop() method for a set randomly extracts and returns an element and pops it out of the set. We make a copy of the set before popping the element so that the original list is not modified.