Related
I'm working with arrays of integer, all of the same size l.
I have a static set of them and I need to build a function to efficiently look them up.
The tricky part is that the elements in the array I need to search might be off by 1.
Given the arrays {A_1, A_2, ..., A_n}, and an array S, I need a function search such that:
search(S)=x iff ∀i: A_x[i] ∈ {S[i]-1, S[i], S[i]+1}.
A possible solution is treating each vector as a point in an l-dimensional space and looking for the closest point, but it'd cost something like O(l*n) in space and O(l*log(n)) in time.
Would there be a solution with a better space complexity (and/or time, of course)?
My arrays are pretty different from each other, and good heuristics might be enough.
Consider a search array S with the values:
S = [s1, s2, s3, ... , sl]
and the average value:
s̅ = (s1 + s2 + s3 + ... + sl) / l
and two matching arrays, one where every value is one greater than the corresponding value in S, and one where very value is one smaller:
A1 = [s1+1, s2+1, s3+1, ... , sl+1]
A2 = [s1−1, s2−1, s3−1, ... , sl−1]
These two arrays would have the average values:
a̅1 = (s1 + 1 + s2 + 1 + s3 + 1 + ... + sl + 1) / l = s̅ + 1
a̅2 = (s1 − 1 + s2 − 1 + s3 − 1 + ... + sl − 1) / l = s̅ − 1
So every matching array, whose values are at most 1 away from the corresponding values in the search array, has an average value that is at most 1 away from the average value of the search array.
If you calculate and store the average value of each array, and then sort the arrays based on their average value (or use an extra data structure that enables you to find all arrays with a certain average value), you can quickly identify which arrays have an average value within 1 of the search array's average value. Depending on the data, this could drastically reduce the number of arrays you have to check for similarity.
After having pre-processed the arrays and stores their average values, performing a search would mean iterating over the search array to calculate the average value, looking up which arrays have a similar average value, and then iterating over those arrays to check every value.
If you expect many arrays to have a similar average value, you could use several averages to detect arrays that are locally very different but similar on average. You could e.g. calculate these four averages:
the first half of the array
the second half of the array
the odd-numbered elements
the even-numbered elements
Analysis of the actual data should give you more information about how to divide the array and combine different averages to be most effective.
If the total sum of an array cannot exceed the integer size, you could store the total sum of each array, and check whether it is within l of the total sum of the search array, instead of using averages. This would avoid having to use floats and divisions.
(You could expand this idea by also storing other properties which are easily calculated and don't take up much space to store, such as the highest and lowest value, the biggest jump, ... They could help create a fingerprint of each array that is near-unique, depending on the data.)
If the number of dimensions is not very small, then probably the best solution will be to build a decision tree that recursively partitions the set along different dimensions.
Each node, including the root, would be a hash table from the possible values for some dimension to either:
The list of points that match that value within tolerance, if it's small enough; or
Those same points in a similar tree partitioning on the remaining dimensions.
Since each level completely eliminates one dimension, the depth of the tree is at most L, and search takes O(L) time.
The order in which the dimensions are chosen along each path is important, of course -- the wrong choice could explode the size of the data structure, with each point appearing many times.
Since your points are "pretty different", though, it should be possible to build a tree with minimal duplication. I would try the ID3 algorithm to choose the dimensions: https://en.wikipedia.org/wiki/ID3_algorithm. That basically means you greedily choose the dimension that maximizes the overall reduction in set size, using an entropy metric.
I would personally create something like a Trie for the lookup. I said "something like" because we have up to 3 values per index that might match. So we aren't creating a decision tree, but a DAG. Where sometimes we have choices.
That is straightforward and will run (with backtracking) in maximum time O(k*l).
But here is the trick. Whenever we see a choice of matching states that we can go into next, we can create a merged state which tries all of them. We can create a few or a lot of these merged states. Each one will defer a choice by 1 step. And if we're careful to keep track of which merged states we've created, we can reuse the same one over and over again.
In theory we can be generating partial matches for somewhat arbitrary subsets of our arrays. Which can grow exponentially in the number of arrays. In practice are likely to only wind up with a few of these merged states. But still we can guarantee a tradeoff - more states up front runs faster later. So we optimize until we are done or have hit the limit of how much data we want to have.
Here is some proof of concept code for this in Python. It will likely build the matcher in time O(n*l) and match in time O(l). However it is only guaranteed to build the matcher in time O(n^2 * l^2) and match in time O(n * l).
import pprint
class Matcher:
def __init__ (self, arrays, optimize_limit=None):
# These are the partial states we could be in during a match.
self.states = [{}]
# By state, this is what we would be trying to match.
self.state_for = ['start']
# By combination we could try to match for, which state it is.
self.comb_state = {'start': 0}
for i in range(len(arrays)):
arr = arrays[i]
# Set up "matched the end".
state_index = len(self.states)
this_state = {'matched': [i]}
self.comb_state[(i, len(arr))] = state_index
self.states.append(this_state)
self.state_for.append((i, len(arr)))
for j in reversed(range(len(arr))):
this_for = (i, j)
prev_state = {}
if 0 == j:
prev_state = self.states[0]
matching_values = set((arr[k] for k in range(max(j-1, 0), min(j+2, len(arr)))))
for v in matching_values:
if v in prev_state:
prev_state[v].append(state_index)
else:
prev_state[v] = [state_index]
if 0 < j:
state_index = len(self.states)
self.states.append(prev_state)
self.state_for.append(this_for)
self.comb_state[this_for] = state_index
# Theoretically optimization can take space
# O(2**len(arrays) * len(arrays[0]))
# We will optimize until we are done or hit a more reasonable limit.
if optimize_limit is None:
# Normally
optimize_limit = len(self.states)**2
# First we find all of the choices at the root.
# This will be an array of arrays with format:
# [state, key, values]
todo = []
for k, v in self.states[0].iteritems():
if 1 < len(v):
todo.append([self.states[0], k, tuple(v)])
while len(todo) and len(self.states) < optimize_limit:
this_state, this_key, this_match = todo.pop(0)
if this_key == 'matched':
pass # We do not need to optimize this!
elif this_match in self.comb_state:
this_state[this_key] = self.comb_state[this_match]
else:
# Construct a new state that is all of these.
new_state = {}
for state_ind in this_match:
for k, v in self.states[state_ind].iteritems():
if k in new_state:
new_state[k] = new_state[k] + v
else:
new_state[k] = v
i = len(self.states)
self.states.append(new_state)
self.comb_state[this_match] = i
self.state_for.append(this_match)
this_state[this_key] = [i]
for k, v in new_state.iteritems():
if 1 < len(v):
todo.append([new_state, k, tuple(v)])
#pp = pprint.PrettyPrinter()
#pp.pprint(self.states)
#pp.pprint(self.comb_state)
#pp.pprint(self.state_for)
def match (self, list1, ind=0, state=0):
this_state = self.states[state]
if 'matched' in this_state:
return this_state['matched']
elif list1[ind] in this_state:
answer = []
for next_state in this_state[list1[ind]]:
answer = answer + self.match(list1, ind+1, next_state)
return answer;
else:
return []
foo = Matcher([[1, 2, 3], [2, 3, 4]])
print(foo.match([2, 2, 3]))
Please note that I deliberately set up a situation where there are 2 matches. It reports both of them. :-)
I came up with a further approach derived off Matt Timmermans's answer: building a simple decision tree that might have certain some arrays in multiple branches. It works even if the error in the array I'm searching is larger than 1.
The idea is the following: given the set of arrays As...
Pick an index and a pivot.
I fixed the pivot to a constant value that works well with my data, and tried all indices to find the best one. Trying multiple pivots might work better, but I didn't need to.
Partition As into two possibly-intersecting subsets, one for the arrays (whose index-th element is) smaller than the pivot, one for the larger arrays. Arrays very close to the pivot are added to both sets:
function partition( As, pivot, index ):
return {
As.filter( A => A[index] <= pivot + 1 ),
As.filter( A => A[index] >= pivot - 1 ),
}
Apply both previous steps to each subset recursively, stopping when a subset only contains a single element.
Here an example of a possible tree generated with this algorithm (note that A2 appears both on the left and right child of the root node):
{A1, A2, A3, A4}
pivot:15
index:73
/ \
/ \
{A1, A2} {A2, A3, A4}
pivot:7 pivot:33
index:54 index:0
/ \ / \
/ \ / \
A1 A2 {A2, A3} A4
pivot:5
index:48
/ \
/ \
A2 A3
The search function then uses this as a normal decision tree: it starts from the root node and recurses either to the left or the right child depending on whether its value at index currentNode.index is greater or less than currentNode.pivot. It proceeds recursively until it reaches a leaf.
Once the decision tree is built, the time complexity is in the worst case O(n), but in practice it's probably closer to O(log(n)) if we choose good indices and pivots (and if the dataset is diverse enough) and find a fairly balanced tree.
The space complexity can be really bad in the worst case (O(2^n)), but it's closer to O(n) with balanced trees.
I'm trying to find a name for my problem, so I don't have to re-invent wheel when coding an algorithm which solves it...
I have say 2,000 binary (row) vectors and I need to pick 500 from them. In the picked sample I do column sums and I want my sample to be as close as possible to a pre-defined distribution of the column sums. I'll be working with 20 to 60 columns.
A tiny example:
Out of the vectors:
110
010
011
110
100
I need to pick 2 to get column sums 2, 1, 0. The solution (exact in this case) would be
110
100
My ideas so far
one could maybe call this a binary multidimensional knapsack, but I did not find any algos for that
Linear Programming could help, but I'd need some step by step explanation as I got no experience with it
as exact solution is not always feasible, something like simulated annealing brute force could work well
a hacky way using constraint solvers comes to mind - first set the constraints tight and gradually loosen them until some solution is found - given that CSP should be much faster than ILP...?
My concrete, practical (if the approximation guarantee works out for you) suggestion would be to apply the maximum entropy method (in Chapter 7 of Boyd and Vandenberghe's book Convex Optimization; you can probably find several implementations with your favorite search engine) to find the maximum entropy probability distribution on row indexes such that (1) no row index is more likely than 1/500 (2) the expected value of the row vector chosen is 1/500th of the predefined distribution. Given this distribution, choose each row independently with probability 500 times its distribution likelihood, which will give you 500 rows on average. If you need exactly 500, repeat until you get exactly 500 (shouldn't take too many tries due to concentration bounds).
Firstly I will make some assumptions regarding this problem:
Regardless whether the column sum of the selected solution is over or under the target, it weighs the same.
The sum of the first, second, and third column are equally weighted in the solution (i.e. If there's a solution whereas the first column sum is off by 1, and another where the third column sum is off by 1, the solution are equally good).
The closest problem I can think of this problem is the Subset sum problem, which itself can be thought of a special case of Knapsack problem.
However both of these problem are NP-Complete. This means there are no polynomial time algorithm that can solve them, even though it is easy to verify the solution.
If I were you the two most arguably efficient solution of this problem are linear programming and machine learning.
Depending on how many columns you are optimising in this problem, with linear programming you can control how much finely tuned you want the solution, in exchange of time. You should read up on this, because this is fairly simple and efficient.
With Machine learning, you need a lot of data sets (the set of vectors and the set of solutions). You don't even need to specify what you want, a lot of machine learning algorithms can generally deduce what you want them to optimise based on your data set.
Both solution has pros and cons, you should decide which one to use yourself based on the circumstances and problem set.
This definitely can be modeled as (integer!) linear program (many problems can). Once you have it, you can use a program such as lpsolve to solve it.
We model vector i is selected as x_i which can be 0 or 1.
Then for each column c, we have a constraint:
sum of all (x_i * value of i in column c) = target for column c
Taking your example, in lp_solve this could look like:
min: ;
+x1 +x4 +x5 >= 2;
+x1 +x4 +x5 <= 2;
+x1 +x2 +x3 +x4 <= 1;
+x1 +x2 +x3 +x4 >= 1;
+x3 <= 0;
+x3 >= 0;
bin x1, x2, x3, x4, x5;
If you are fine with a heuristic based search approach, here is one.
Go over the list and find the minimum squared sum of the digit wise difference between each bit string and the goal. For example, if we are looking for 2, 1, 0, and we are scoring 0, 1, 0, we would do it in the following way:
Take the digit wise difference:
2, 0, 1
Square the digit wise difference:
4, 0, 1
Sum:
5
As a side note, squaring the difference when scoring is a common method when doing heuristic search. In your case, it makes sense because bit strings that have a 1 in as the first digit are a lot more interesting to us. In your case this simple algorithm would pick first 110, then 100, which would is the best solution.
In any case, there are some optimizations that could be made to this, I will post them here if this kind of approach is what you are looking for, but this is the core of the algorithm.
You have a given target binary vector. You want to select M vectors out of N that have the closest sum to the target. Let's say you use the eucilidean distance to measure if a selection is better than another.
If you want an exact sum, have a look at the k-sum problem which is a generalization of the 3SUM problem. The problem is harder than the subset sum problem, because you want an exact number of elements to add to a target value. There is a solution in O(N^(M/2)). lg N), but that means more than 2000^250 * 7.6 > 10^826 operations in your case (in the favorable case where vectors operations have a cost of 1).
First conclusion: do not try to get an exact result unless your vectors have some characteristics that may reduce the complexity.
Here's a hill climbing approach:
sort the vectors by number of 1's: 111... first, 000... last;
use the polynomial time approximate algorithm for the subset sum;
you have an approximate solution with K elements. Because of the order of elements (the big ones come first), K should be a little as possible:
if K >= M, you take the M first vectors of the solution and that's probably near the best you can do.
if K < M, you can remove the first vector and try to replace it with 2 or more vectors from the rest of the N vectors, using the same technique, until you have M vectors. To sumarize: split the big vectors into smaller ones until you reach the correct number of vectors.
Here's a proof of concept with numbers, in Python:
import random
def distance(x, y):
return abs(x-y)
def show(ls):
if len(ls) < 10:
return str(ls)
else:
return ", ".join(map(str, ls[:5]+("...",)+ls[-5:]))
def find(is_xs, target):
# see https://en.wikipedia.org/wiki/Subset_sum_problem#Pseudo-polynomial_time_dynamic_programming_solution
S = [(0, ())] # we store indices along with values to get the path
for i, x in is_xs:
T = [(x + t, js + (i,)) for t, js in S]
U = sorted(S + T)
y, ks = U[0]
S = [(y, ks)]
for z, ls in U:
if z == target: # use the euclidean distance here if you want an approximation
return ls
if z != y and z < target:
y, ks = z, ls
S.append((z, ls))
ls = S[-1][1] # take the closest element to target
return ls
N = 2000
M = 500
target = 1000
xs = [random.randint(0, 10) for _ in range(N)]
print ("Take {} numbers out of {} to make a sum of {}", M, xs, target)
xs = sorted(xs, reverse = True)
is_xs = list(enumerate(xs))
print ("Sorted numbers: {}".format(show(tuple(is_xs))))
ls = find(is_xs, target)
print("FIRST TRY: {} elements ({}) -> {}".format(len(ls), show(ls), sum(x for i, x in is_xs if i in ls)))
splits = 0
while len(ls) < M:
first_x = xs[ls[0]]
js_ys = [(i, x) for i, x in is_xs if i not in ls and x != first_x]
replace = find(js_ys, first_x)
splits += 1
if len(replace) < 2 or len(replace) + len(ls) - 1 > M or sum(xs[i] for i in replace) != first_x:
print("Give up: can't replace {}.\nAdd the lowest elements.")
ls += tuple([i for i, x in is_xs if i not in ls][len(ls)-M:])
break
print ("Replace {} (={}) by {} (={})".format(ls[:1], first_x, replace, sum(xs[i] for i in replace)))
ls = tuple(sorted(ls[1:] + replace)) # use a heap?
print("{} elements ({}) -> {}".format(len(ls), show(ls), sum(x for i, x in is_xs if i in ls)))
print("AFTER {} splits, {} -> {}".format(splits, ls, sum(x for i, x in is_xs if i in ls)))
The result is obviously not guaranteed to be optimal.
Remarks:
Complexity: find has a polynomial time complexity (see the Wikipedia page) and is called at most M^2 times, hence the complexity remains polynomial. In practice, the process is reasonably fast (split calls have a small target).
Vectors: to ensure that you reach the target with the minimum of elements, you can improve the order of element. Your target is (t_1, ..., t_c): if you sort the t_js from max to min, you get the more importants columns first. You can sort the vectors: by number of 1s and then by the presence of a 1 in the most important columns. E.g. target = 4 8 6 => 1 1 1 > 0 1 1 > 1 1 0 > 1 0 1 > 0 1 0 > 0 0 1 > 1 0 0 > 0 0 0.
find (Vectors) if the current sum exceed the target in all the columns, then you're not connecting to the target (any vector you add to the current sum will bring you farther from the target): don't add the sum to S (z >= target case for numbers).
I propose a simple ad hoc algorithm, which, broadly speaking, is a kind of gradient descent algorithm. It seems to work relatively well for input vectors which have a distribution of 1s “similar” to the target sum vector, and probably also for all “nice” input vectors, as defined in a comment of yours. The solution is not exact, but the approximation seems good.
The distance between the sum vector of the output vectors and the target vector is taken to be Euclidean. To minimize it means minimizing the sum of the square differences off sum vector and target vector (the square root is not needed because it is monotonic). The algorithm does not guarantee to yield the sample that minimizes the distance from the target, but anyway makes a serious attempt at doing so, by always moving in some locally optimal direction.
The algorithm can be split into 3 parts.
First of all the first M candidate output vectors out of the N input vectors (e.g., N=2000, M=500) are put in a list, and the remaining vectors are put in another.
Then "approximately optimal" swaps between vectors in the two lists are done, until either the distance would not decrease any more, or a predefined maximum number of iterations is reached. An approximately optimal swap is one where removing the first vector from the list of output vectors causes a maximal decrease or minimal increase of the distance, and then, after the removal of the first vector, adding the second vector to the same list causes a maximal decrease of the distance. The whole swap is avoided if the net result is not a decrease of the distance.
Then, as a last phase, "optimal" swaps are done, again stopping on no decrease in distance or maximum number of iterations reached. Optimal swaps cause a maximal decrease of the distance, without requiring the removal of the first vector to be optimal in itself. To find an optimal swap all vector pairs have to be checked. This phase is much more expensive, being O(M(N-M)), while the previous "approximate" phase is O(M+(N-M))=O(N). Luckily, when entering this phase, most of the work has already been done by the previous phase.
from typing import List, Tuple
def get_sample(vects: List[Tuple[int]], target: Tuple[int], n_out: int,
max_approx_swaps: int = None, max_optimal_swaps: int = None,
verbose: bool = False) -> List[Tuple[int]]:
"""
Get a sample of the input vectors having a sum close to the target vector.
Closeness is measured in Euclidean metrics. The output is not guaranteed to be
optimal (minimum square distance from target), but a serious attempt is made.
The max_* parameters can be used to avoid too long execution times,
tune them to your needs by setting verbose to True, or leave them None (∞).
:param vects: the list of vectors (tuples) with the same number of "columns"
:param target: the target vector, with the same number of "columns"
:param n_out: the requested sample size
:param max_approx_swaps: the max number of approximately optimal vector swaps,
None means unlimited (default: None)
:param max_optimal_swaps: the max number of optimal vector swaps,
None means unlimited (default: None)
:param verbose: print some info if True (default: False)
:return: the sample of n_out vectors having a sum close to the target vector
"""
def square_distance(v1, v2):
return sum((e1 - e2) ** 2 for e1, e2 in zip(v1, v2))
n_vec = len(vects)
assert n_vec > 0
assert n_out > 0
n_rem = n_vec - n_out
assert n_rem > 0
output = vects[:n_out]
remain = vects[n_out:]
n_col = len(vects[0])
assert n_col == len(target) > 0
sumvect = (0,) * n_col
for outvect in output:
sumvect = tuple(map(int.__add__, sumvect, outvect))
sqdist = square_distance(sumvect, target)
if verbose:
print(f"sqdist = {sqdist:4} after"
f" picking the first {n_out} vectors out of {n_vec}")
if max_approx_swaps is None:
max_approx_swaps = sqdist
n_approx_swaps = 0
while sqdist and n_approx_swaps < max_approx_swaps:
# find the best vect to subtract (the square distance MAY increase)
sqdist_0 = None
index_0 = None
sumvect_0 = None
for index in range(n_out):
tmp_sumvect = tuple(map(int.__sub__, sumvect, output[index]))
tmp_sqdist = square_distance(tmp_sumvect, target)
if sqdist_0 is None or sqdist_0 > tmp_sqdist:
sqdist_0 = tmp_sqdist
index_0 = index
sumvect_0 = tmp_sumvect
# find the best vect to add,
# but only if there is a net decrease of the square distance
sqdist_1 = sqdist
index_1 = None
sumvect_1 = None
for index in range(n_rem):
tmp_sumvect = tuple(map(int.__add__, sumvect_0, remain[index]))
tmp_sqdist = square_distance(tmp_sumvect, target)
if sqdist_1 > tmp_sqdist:
sqdist_1 = tmp_sqdist
index_1 = index
sumvect_1 = tmp_sumvect
if sumvect_1:
tmp = output[index_0]
output[index_0] = remain[index_1]
remain[index_1] = tmp
sqdist = sqdist_1
sumvect = sumvect_1
n_approx_swaps += 1
else:
break
if verbose:
print(f"sqdist = {sqdist:4} after {n_approx_swaps}"
f" approximately optimal swap{'s'[n_approx_swaps == 1:]}")
diffvect = tuple(map(int.__sub__, sumvect, target))
if max_optimal_swaps is None:
max_optimal_swaps = sqdist
n_optimal_swaps = 0
while sqdist and n_optimal_swaps < max_optimal_swaps:
# find the best pair to swap,
# but only if the square distance decreases
best_sqdist = sqdist
best_diffvect = diffvect
best_pair = None
for i0 in range(M):
tmp_diffvect = tuple(map(int.__sub__, diffvect, output[i0]))
for i1 in range(n_rem):
new_diffvect = tuple(map(int.__add__, tmp_diffvect, remain[i1]))
new_sqdist = sum(d * d for d in new_diffvect)
if best_sqdist > new_sqdist:
best_sqdist = new_sqdist
best_diffvect = new_diffvect
best_pair = (i0, i1)
if best_pair:
tmp = output[best_pair[0]]
output[best_pair[0]] = remain[best_pair[1]]
remain[best_pair[1]] = tmp
sqdist = best_sqdist
diffvect = best_diffvect
n_optimal_swaps += 1
else:
break
if verbose:
print(f"sqdist = {sqdist:4} after {n_optimal_swaps}"
f" optimal swap{'s'[n_optimal_swaps == 1:]}")
return output
from random import randrange
C = 30 # number of columns
N = 2000 # total number of vectors
M = 500 # number of output vectors
F = 0.9 # fill factor of the target sum vector
T = int(M * F) # maximum value + 1 that can be appear in the target sum vector
A = 10000 # maximum number of approximately optimal swaps, may be None (∞)
B = 10 # maximum number of optimal swaps, may be None (unlimited)
target = tuple(randrange(T) for _ in range(C))
vects = [tuple(int(randrange(M) < t) for t in target) for _ in range(N)]
sample = get_sample(vects, target, M, A, B, True)
Typical output:
sqdist = 2639 after picking the first 500 vectors out of 2000
sqdist = 9 after 27 approximately optimal swaps
sqdist = 1 after 4 optimal swaps
P.S.: As it stands, this algorithm is not limited to binary input vectors, integer vectors would work too. Intuitively I suspect that the quality of the optimization could suffer, though. I suspect that this algorithm is more appropriate for binary vectors.
P.P.S.: Execution times with your kind of data are probably acceptable with standard CPython, but get better (like a couple of seconds, almost a factor of 10) with PyPy. To handle bigger sets of data, the algorithm would have to be translated to C or some other language, which should not be difficult at all.
Recently I needed to do weighted random selection of elements from a list, both with and without replacement. While there are well known and good algorithms for unweighted selection, and some for weighted selection without replacement (such as modifications of the resevoir algorithm), I couldn't find any good algorithms for weighted selection with replacement. I also wanted to avoid the resevoir method, as I was selecting a significant fraction of the list, which is small enough to hold in memory.
Does anyone have any suggestions on the best approach in this situation? I have my own solutions, but I'm hoping to find something more efficient, simpler, or both.
One of the fastest ways to make many with replacement samples from an unchanging list is the alias method. The core intuition is that we can create a set of equal-sized bins for the weighted list that can be indexed very efficiently through bit operations, to avoid a binary search. It will turn out that, done correctly, we will need to only store two items from the original list per bin, and thus can represent the split with a single percentage.
Let's us take the example of five equally weighted choices, (a:1, b:1, c:1, d:1, e:1)
To create the alias lookup:
Normalize the weights such that they sum to 1.0. (a:0.2 b:0.2 c:0.2 d:0.2 e:0.2) This is the probability of choosing each weight.
Find the smallest power of 2 greater than or equal to the number of variables, and create this number of partitions, |p|. Each partition represents a probability mass of 1/|p|. In this case, we create 8 partitions, each able to contain 0.125.
Take the variable with the least remaining weight, and place as much of it's mass as possible in an empty partition. In this example, we see that a fills the first partition. (p1{a|null,1.0},p2,p3,p4,p5,p6,p7,p8) with (a:0.075, b:0.2 c:0.2 d:0.2 e:0.2)
If the partition is not filled, take the variable with the most weight, and fill the partition with that variable.
Repeat steps 3 and 4, until none of the weight from the original partition need be assigned to the list.
For example, if we run another iteration of 3 and 4, we see
(p1{a|null,1.0},p2{a|b,0.6},p3,p4,p5,p6,p7,p8) with (a:0, b:0.15 c:0.2 d:0.2 e:0.2) left to be assigned
At runtime:
Get a U(0,1) random number, say binary 0.001100000
bitshift it lg2(p), finding the index partition. Thus, we shift it by 3, yielding 001.1, or position 1, and thus partition 2.
If the partition is split, use the decimal portion of the shifted random number to decide the split. In this case, the value is 0.5, and 0.5 < 0.6, so return a.
Here is some code and another explanation, but unfortunately it doesn't use the bitshifting technique, nor have I actually verified it.
A simple approach that hasn't been mentioned here is one proposed in Efraimidis and Spirakis. In python you could select m items from n >= m weighted items with strictly positive weights stored in weights, returning the selected indices, with:
import heapq
import math
import random
def WeightedSelectionWithoutReplacement(weights, m):
elt = [(math.log(random.random()) / weights[i], i) for i in range(len(weights))]
return [x[1] for x in heapq.nlargest(m, elt)]
This is very similar in structure to the first approach proposed by Nick Johnson. Unfortunately, that approach is biased in selecting the elements (see the comments on the method). Efraimidis and Spirakis proved that their approach is equivalent to random sampling without replacement in the linked paper.
Here's what I came up with for weighted selection without replacement:
def WeightedSelectionWithoutReplacement(l, n):
"""Selects without replacement n random elements from a list of (weight, item) tuples."""
l = sorted((random.random() * x[0], x[1]) for x in l)
return l[-n:]
This is O(m log m) on the number of items in the list to be selected from. I'm fairly certain this will weight items correctly, though I haven't verified it in any formal sense.
Here's what I came up with for weighted selection with replacement:
def WeightedSelectionWithReplacement(l, n):
"""Selects with replacement n random elements from a list of (weight, item) tuples."""
cuml = []
total_weight = 0.0
for weight, item in l:
total_weight += weight
cuml.append((total_weight, item))
return [cuml[bisect.bisect(cuml, random.random()*total_weight)] for x in range(n)]
This is O(m + n log m), where m is the number of items in the input list, and n is the number of items to be selected.
I'd recommend you start by looking at section 3.4.2 of Donald Knuth's Seminumerical Algorithms.
If your arrays are large, there are more efficient algorithms in chapter 3 of Principles of Random Variate Generation by John Dagpunar. If your arrays are not terribly large or you're not concerned with squeezing out as much efficiency as possible, the simpler algorithms in Knuth are probably fine.
It is possible to do Weighted Random Selection with replacement in O(1) time, after first creating an additional O(N)-sized data structure in O(N) time. The algorithm is based on the Alias Method developed by Walker and Vose, which is well described here.
The essential idea is that each bin in a histogram would be chosen with probability 1/N by a uniform RNG. So we will walk through it, and for any underpopulated bin which would would receive excess hits, assign the excess to an overpopulated bin. For each bin, we store the percentage of hits which belong to it, and the partner bin for the excess. This version tracks small and large bins in place, removing the need for an additional stack. It uses the index of the partner (stored in bucket[1]) as an indicator that they have already been processed.
Here is a minimal python implementation, based on the C implementation here
def prep(weights):
data_sz = len(weights)
factor = data_sz/float(sum(weights))
data = [[w*factor, i] for i,w in enumerate(weights)]
big=0
while big<data_sz and data[big][0]<=1.0: big+=1
for small,bucket in enumerate(data):
if bucket[1] is not small: continue
excess = 1.0 - bucket[0]
while excess > 0:
if big==data_sz: break
bucket[1] = big
bucket = data[big]
bucket[0] -= excess
excess = 1.0 - bucket[0]
if (excess >= 0):
big+=1
while big<data_sz and data[big][0]<=1: big+=1
return data
def sample(data):
r=random.random()*len(data)
idx = int(r)
return data[idx][1] if r-idx > data[idx][0] else idx
Example usage:
TRIALS=1000
weights = [20,1.5,9.8,10,15,10,15.5,10,8,.2];
samples = [0]*len(weights)
data = prep(weights)
for _ in range(int(sum(weights)*TRIALS)):
samples[sample(data)]+=1
result = [float(s)/TRIALS for s in samples]
err = [a-b for a,b in zip(result,weights)]
print(result)
print([round(e,5) for e in err])
print(sum([e*e for e in err]))
The following is a description of random weighted selection of an element of a
set (or multiset, if repeats are allowed), both with and without replacement in O(n) space
and O(log n) time.
It consists of implementing a binary search tree, sorted by the elements to be
selected, where each node of the tree contains:
the element itself (element)
the un-normalized weight of the element (elementweight), and
the sum of all the un-normalized weights of the left-child node and all of
its children (leftbranchweight).
the sum of all the un-normalized weights of the right-child node and all of
its chilren (rightbranchweight).
Then we randomly select an element from the BST by descending down the tree. A
rough description of the algorithm follows. The algorithm is given a node of
the tree. Then the values of leftbranchweight, rightbranchweight,
and elementweight of node is summed, and the weights are divided by this
sum, resulting in the values leftbranchprobability,
rightbranchprobability, and elementprobability, respectively. Then a
random number between 0 and 1 (randomnumber) is obtained.
if the number is less than elementprobability,
remove the element from the BST as normal, updating leftbranchweight
and rightbranchweight of all the necessary nodes, and return the
element.
else if the number is less than (elementprobability + leftbranchweight)
recurse on leftchild (run the algorithm using leftchild as node)
else
recurse on rightchild
When we finally find, using these weights, which element is to be returned, we either simply return it (with replacement) or we remove it and update relevant weights in the tree (without replacement).
DISCLAIMER: The algorithm is rough, and a treatise on the proper implementation
of a BST is not attempted here; rather, it is hoped that this answer will help
those who really need fast weighted selection without replacement (like I do).
This is an old question for which numpy now offers an easy solution so I thought I would mention it. Current version of numpy is version 1.2 and numpy.random.choice allows the sampling to be done with or without replacement and with given weights.
Suppose you want to sample 3 elements without replacement from the list ['white','blue','black','yellow','green'] with a prob. distribution [0.1, 0.2, 0.4, 0.1, 0.2]. Using numpy.random module it is as easy as this:
import numpy.random as rnd
sampling_size = 3
domain = ['white','blue','black','yellow','green']
probs = [.1, .2, .4, .1, .2]
sample = rnd.choice(domain, size=sampling_size, replace=False, p=probs)
# in short: rnd.choice(domain, sampling_size, False, probs)
print(sample)
# Possible output: ['white' 'black' 'blue']
Setting the replace flag to True, you have a sampling with replacement.
More info here:
http://docs.scipy.org/doc/numpy/reference/generated/numpy.random.choice.html#numpy.random.choice
We faced a problem to randomly select K validators of N candidates once per epoch proportionally to their stakes. But this gives us the following problem:
Imagine probabilities of each candidate:
0.1
0.1
0.8
Probabilities of each candidate after 1'000'000 selections 2 of 3 without replacement became:
0.254315
0.256755
0.488930
You should know, those original probabilities are not achievable for 2 of 3 selection without replacement.
But we wish initial probabilities to be a profit distribution probabilities. Else it makes small candidate pools more profitable. So we realized that random selection with replacement would help us – to randomly select >K of N and store also weight of each validator for reward distribution:
std::vector<int> validators;
std::vector<int> weights(n);
int totalWeights = 0;
for (int j = 0; validators.size() < m; j++) {
int value = rand() % likehoodsSum;
for (int i = 0; i < n; i++) {
if (value < likehoods[i]) {
if (weights[i] == 0) {
validators.push_back(i);
}
weights[i]++;
totalWeights++;
break;
}
value -= likehoods[i];
}
}
It gives an almost original distribution of rewards on millions of samples:
0.101230
0.099113
0.799657
I want to implement an iterative algorithm, which calculates weighted average. The specific weight law does not matter, but it should be close to 1 for the newest values and close to 0 to the oldest.
The algorithm should be iterative. i.e. it should not remember all previous values. It should know only one newest value and any aggregative information about past, like previous values of the average, sums, counts etc.
Is it possible?
For example, the following algorithm can be:
void iterate(double value) {
sum *= 0.99;
sum += value;
count++;
avg = sum / count;
}
It will give exponential decreasing weight, which may be not good. Is it possible to have step decreasing weight or something?
EDIT 1
The the requirements for weighing law is follows:
1) The weight decreases into past
2) I has some mean or characteristic duration so that values older this duration matters much lesser than newer ones
3) I should be able to set this duration
EDIT 2
I need the following. Suppose v_i are values, where v_1 is the first. Also suppose w_i are weights. But w_0 is THE LAST.
So, after first value came I have first average
a_1 = v_1 * w_0
After the second value v_2 came, I should have average
a_2 = v_1 * w_1 + v_2 * w_0
With next value I should have
a_3 = v_1 * w_2 + v_2 * w_1 + v_3 * w_0
Note, that weight profile is moving with me, while I am moving along value sequence.
I.e. each value does not have it's own weight all the time. My goal is to have this weight lower while going to past.
First a bit of background. If we were keeping a normal average, it would go like this:
average(a) = 11
average(a,b) = (average(a)+b)/2
average(a,b,c) = (average(a,b)*2 + c)/3
average(a,b,c,d) = (average(a,b,c)*3 + d)/4
As you can see here, this is an "online" algorithm and we only need to keep track of pieces of data: 1) the total numbers in the average, and 2) the average itself. Then we can undivide the average by the total, add in the new number, and divide it by the new total.
Weighted averages are a bit different. It depends on what kind of weighted average. For example if you defined:
weightedAverage(a,wa, b,wb, c,wc, ..., z,wz) = a*wa + b*wb + c*wc + ... + w*wz
or
weightedAverage(elements, weights) = elements·weights
...then you don't need to do anything besides add the new element*weight! If however you defined the weighted average akin to an expected-value from probability:
weightedAverage(elements,weights) = elements·weights / sum(weights)
...then you'd need to keep track of the total weights. Instead of undividing by the total number of elements, you undivide by the total weight, add in the new element*weight, then divide by the new total weight.
Alternatively you don't need to undivide, as demonstrated below: you can merely keep track of the temporary dot product and weight total in a closure or an object, and divide it as you yield (this can help a lot with avoiding numerical inaccuracy from compounded rounding errors).
In python this would be:
def makeAverager():
dotProduct = 0
totalWeight = 0
def averager(newValue, weight):
nonlocal dotProduct,totalWeight
dotProduct += newValue*weight
totalWeight += weight
return dotProduct/totalWeight
return averager
Demo:
>>> averager = makeAverager()
>>> [averager(value,w) for value,w in [(100,0.2), (50,0.5), (100,0.1)]]
[100.0, 64.28571428571429, 68.75]
>>> averager(10,1.1)
34.73684210526316
>>> averager(10,1.1)
25.666666666666668
>>> averager(30,2.0)
27.4
> But my task is to have average recalculated each time new value arrives having old values reweighted. –OP
Your task is almost always impossible, even with exceptionally simple weighting schemes.
You are asking to, with O(1) memory, yield averages with a changing weighting scheme. For example, {values·weights1, (values+[newValue2])·weights2, (values+[newValue2,newValue3])·weights3, ...} as new values are being passed in, for some nearly arbitrarily changing weights sequence. This is impossible due to injectivity. Once you merge the numbers in together, you lose a massive amount of information. For example, even if you had the weight vector, you could not recover the original value vector, or vice versa. There are only two cases I can think of where you could get away with this:
Constant weights such as [2,2,2,...2]: this is equivalent to an on-line averaging algorithm, which you don't want because the old values are not being "reweighted".
The relative weights of previous answers do not change. For example you could do weights of [8,4,2,1], and add in a new element with arbitrary weight like ...+[1], but you must increase all the previous by the same multiplicative factor, like [16,8,4,2]+[1]. Thus at each step, you are adding a new arbitrary weight, and a new arbitrary rescaling of the past, so you have 2 degrees of freedom (only 1 if you need to keep your dot-product normalized). The weight-vectors you'd get would look like:
[w0]
[w0*(s1), w1]
[w0*(s1*s2), w1*(s2), w2]
[w0*(s1*s2*s3), w1*(s2*s3), w2*(s3), w3]
...
Thus any weighting scheme you can make look like that will work (unless you need to keep the thing normalized by the sum of weights, in which case you must then divide the new average by the new sum, which you can calculate by keeping only O(1) memory). Merely multiply the previous average by the new s (which will implicitly distribute over the dot-product into the weights), and tack on the new +w*newValue.
I think you are looking for something like this:
void iterate(double value) {
count++;
weight = max(0, 1 - (count / 1000));
avg = ( avg * total_weight * (count - 1) + weight * value) / (total_weight * (count - 1) + weight)
total_weight += weight;
}
Here I'm assuming you want the weights to sum to 1. As long as you can generate a relative weight without it changing in the future, you can end up with a solution which mimics this behavior.
That is, suppose you defined your weights as a sequence {s_0, s_1, s_2, ..., s_n, ...} and defined the input as sequence {i_0, i_1, i_2, ..., i_n}.
Consider the form: sum(s_0*i_0 + s_1*i_1 + s_2*i_2 + ... + s_n*i_n) / sum(s_0 + s_1 + s_2 + ... + s_n). Note that it is trivially possible to compute this incrementally with a couple of aggregation counters:
int counter = 0;
double numerator = 0;
double denominator = 0;
void addValue(double val)
{
double weight = calculateWeightFromCounter(counter);
numerator += weight * val;
denominator += weight;
}
double getAverage()
{
if (denominator == 0.0) return 0.0;
return numerator / denominator;
}
Of course, calculateWeightFromCounter() in this case shouldn't generate weights that sum to one -- the trick here is that we average by dividing by the sum of the weights so that in the end, the weights virtually seem to sum to one.
The real trick is how you do calculateWeightFromCounter(). You could simply return the counter itself, for example, however note that the last weighted number would not be near the sum of the counters necessarily, so you may not end up with the exact properties you want. (It's hard to say since, as mentioned, you've left a fairly open problem.)
This is too long to post in a comment, but it may be useful to know.
Suppose you have:
w_0*v_n + ... w_n*v_0 (we'll call this w[0..n]*v[n..0] for short)
Then the next step is:
w_0*v_n1 + ... w_n1*v_0 (and this is w[0..n1]*v[n1..0] for short)
This means we need a way to calculate w[1..n1]*v[n..0] from w[0..n]*v[n..0].
It's certainly possible that v[n..0] is 0, ..., 0, z, 0, ..., 0 where z is at some location x.
If we don't have any 'extra' storage, then f(z*w(x))=z*w(x + 1) where w(x) is the weight for location x.
Rearranging the equation, w(x + 1) = f(z*w(x))/z. Well, w(x + 1) better be constant for a constant x, so f(z*w(x))/z better be constant. Hence, f must let z propagate -- that is, f(z*w(x)) = z*f(w(x)).
But here again we have an issue. Note that if z (which could be any number) can propagate through f, then w(x) certainly can. So f(z*w(x)) = w(x)*f(z). Thus f(w(x)) = w(x)/f(z).
But for a constant x, w(x) is constant, and thus f(w(x)) better be constant, too. w(x) is constant, so f(z) better be constant so that w(x)/f(z) is constant. Thus f(w(x)) = w(x)/c where c is a constant.
So, f(x)=c*x where c is a constant when x is a weight value.
So w(x+1) = c*w(x).
That is, each weight is a multiple of the previous. Thus, the weights take the form w(x)=m*b^x.
Note that this assumes the only information f has is the last aggregated value. Note that at some point you will be reduced to this case unless you're willing to store a non-constant amount of data representing your input. You cannot represent an infinite length vector of real numbers with a real number, but you can approximate them somehow in a constant, finite amount of storage. But this would merely be an approximation.
Although I haven't rigorously proven it, it is my conclusion that what you want is impossible to do with a high degree of precision, but you may be able to use log(n) space (which may as well be O(1) for many practical applications) to generate a quality approximation. You may be able to use even less.
I tried to practically code something (in Java). As has been said, your goal is not achievable. You can only count average from some number of last remembered values. If you don't need to be exact, you can approximate the older values. I tried to do it by remembering last 5 values exactly and older values only SUMmed by 5 values, remembering the last 5 SUMs. Then, the complexity is O(2n) for remembering last n+n*n values. This is a very rough approximation.
You can modify the "lastValues" and "lasAggregatedSums" array sizes as you want. See this ascii-art picture trying to display a graph of last values, showing that the first columns (older data) are remembered as aggregated value (not individually), and only the earliest 5 values are remembered individually.
values:
#####
##### ##### #
##### ##### ##### # #
##### ##### ##### ##### ## ##
##### ##### ##### ##### ##### #####
time: --->
Challenge 1: My example doesn't count weights, but I think it shouldn't be problem for you to add weights for the "lastAggregatedSums" appropriately - the only problem is, that if you want lower weights for older values, it would be harder, because the array is rotating, so it is not straightforward to know which weight for which array member. Maybe you can modify the algorithm to always "shift" values in the array instead of rotating? Then adding weights shouldn't be a problem.
Challenge 2: The arrays are initialized with 0 values, and those values are counting to the average from the beginning, even when we haven't receive enough values. If you are running the algorithm for long time, you probably don't bother that it is learning for some time at the beginning. If you do, you can post a modification ;-)
public class AverageCounter {
private float[] lastValues = new float[5];
private float[] lastAggregatedSums = new float[5];
private int valIdx = 0;
private int aggValIdx = 0;
private float avg;
public void add(float value) {
lastValues[valIdx++] = value;
if(valIdx == lastValues.length) {
// count average of last values and save into the aggregated array.
float sum = 0;
for(float v: lastValues) {sum += v;}
lastAggregatedSums[aggValIdx++] = sum;
if(aggValIdx >= lastAggregatedSums.length) {
// rotate aggregated values index
aggValIdx = 0;
}
valIdx = 0;
}
float sum = 0;
for(float v: lastValues) {sum += v;}
for(float v: lastAggregatedSums) {sum += v;}
avg = sum / (lastValues.length + lastAggregatedSums.length * lastValues.length);
}
public float getAvg() {
return avg;
}
}
you can combine (weighted sum) exponential means with different effective window sizes (N) in order to get the desired weights.
Use more exponential means to define your weight profile more detailed.
(more exponential means also means to store and calculate more values, so here is the trade off)
A memoryless solution is to calculate the new average from a weighted combination of the previous average and the new value:
average = (1 - P) * average + P * value
where P is an empirical constant, 0 <= P <= 1
expanding gives:
average = sum i (weight[i] * value[i])
where value[0] is the newest value, and
weight[i] = P * (1 - P) ^ i
When P is low, historical values are given higher weighting.
The closer P gets to 1, the more quickly it converges to newer values.
When P = 1, it's a regular assignment and ignores previous values.
If you want to maximise the contribution of value[N], maximize
weight[N] = P * (1 - P) ^ N
where 0 <= P <= 1
I discovered weight[N] is maximized when
P = 1 / (N + 1)
Here is a real-world combinatorial optimization problem.
We are given a large set of value propositions for a certain product. The value propositions are of different types but each type is independent and adds equal benefit to the overall product. In building the product, we can include any non-negative integer number of "units" of each type. However, after adding the first unit of a certain type, the marginal benefit of additional units of that type continually decreases. In fact, the marginal benefit of a new unit is the inverse of the number of units of that type, after adding the new unit. Our product must have a least one unit of some type, and there is a small correction that we must make to the overall value because of this requirement.
Let T[] be an array representing the number of each type in a certain production run of the product. Then the overall value V is given by (pseudo code):
V = 1
For Each t in T
V = V * (t + 1)
Next t
V = V - 1 // correction
On cost side, units of the same type have the same cost. But units of different types each have unique, irrational costs. The number of types is large, but we are given an array of type costs C[] that is sorted from smallest to largest. Let's further assume that the type quantity array T[] is also sorted by cost from smallest to largest. Then the overall cost U is simply the sum of each unit cost:
U = 0
For i = 0, i < NumOfValueTypes
U = U + T[i] * C[i]
Next i
So far so good. So here is the problem: Given product P with value V and cost U, find the product Q with the cost U' and value V', having the minimal U' such that U' > U, V'/U' > V/U.
The problem you've described is nonlinear integer programming problem because it contains a product of integer variables t. Its feasibility set is not closed because of strict inequalities which can be worked around by using non-strict inequalities and adding a small positive number (epsilon) to the right hand sides. Then the problem can be formulated in AMPL as follows:
set Types;
param Costs{Types}; # C
param GivenProductValue; # V
param GivenProductCost; # U
param Epsilon;
var units{Types} integer >= 0; # T
var productCost = sum {t in Types} units[t] * Costs[t];
minimize cost: productCost;
s.t. greaterCost: productCost >= GivenProductCost + Epsilon;
s.t. greaterValuePerCost:
prod {t in Types} (units[t] + 1) - 1 >=
productCost * GivenProductValue / GivenProductCost + Epsilon;
This problem can be solved using a nonlinear integer programming solver such as Couenne.
Honestly I don't think there is an easy way to solve this. The best thing would be to write the system and solve it with a solver ( Excel solver will do the tricks, but you can use Ampl to solve this non lienar program.)
The Program:
Define: U;
V;
C=[c1,...cn];
Variables: T=[t1,t2,...tn];
Objective Function: SUM(ti.ci)
Constraints:
For all i: ti integer
SUM(ti.ci) > U
(PROD(ti+1)-1).U > V.SUM(ti.ci)
It works well with excel, (you just replace >U by >=U+d where d is the significative number of the costs- (i.e if C=[1.1, 1.8, 3.0, 9.3] d =0.1) since excel doesn't allow stric inequalities in the solver.)
I guess with a real solver like Ampl it will work perfectly.
Hope it helps,