Random Pick with Weight
Random Pick with Weight:
You are given a 0-indexed array of positive integers w
where w[i]
describes the weight of the ith
index.
You need to implement the function pickIndex()
, which randomly picks an index in the range [0, w.length - 1]
(inclusive) and returns it. The probability of picking an index i
is w[i] / sum(w)
.
For example, if w = [1, 3]
, the probability of picking index 0
is 1 / (1 + 3) = 0.25
(i.e., 25%
), and the probability of picking index 1
is 3 / (1 + 3) = 0.75
(i.e., 75%
).
Example 1:
Input
["Solution","pickIndex"]
[[[1]],[]]
Output
[null,0]
Explanation
Solution solution = new Solution([1]);
solution.pickIndex(); // return 0. The only option is to
return 0 since there is only one element in w.
Example 2:
Input
["Solution","pickIndex","pickIndex","pickIndex","pickIndex","pickIndex"]
[[[1,3]],[],[],[],[],[]]
Output
[null,1,1,1,1,0]
Explanation
Solution solution = new Solution([1, 3]);
solution.pickIndex(); // return 1. It is returning the
second element (index = 1) that has a probability of 3/4.
solution.pickIndex(); // return 1
solution.pickIndex(); // return 1
solution.pickIndex(); // return 1
solution.pickIndex(); // return 0. It is returning the
first element (index = 0) that has a probability of 1/4.
Since this is a randomization problem, multiple answers are allowed.
All of the following outputs can be considered correct:
[null,1,1,1,1,0]
[null,1,1,1,1,1]
[null,1,1,1,0,0]
[null,1,1,1,0,1]
[null,1,0,1,0,0]
......
and so on.
Constraints:
1 <= w.length <= 10^4
1 <= w[i] <= 10^5
pickIndex
will be called at most10^4
times.
Try this Problem on your own or check similar problems:
Solution:
- Java
- JavaScript
- Python
- C++
class Solution {
Random random;
int[] sums;
public Solution(int[] w) {
random = new Random();
for(int i = 1; i < w.length; ++i){
w[i] += w[i - 1];
}
sums = w;
}
public int pickIndex() {
int idx = random.nextInt(sums[sums.length-1]) + 1;
int i = Arrays.binarySearch(sums, idx);
return i >= 0 ? i : ~i;
}
}
/**
* Your Solution object will be instantiated and called as such:
* Solution obj = new Solution(w);
* int param_1 = obj.pickIndex();
*/
/**
* @param {number[]} w
*/
var Solution = function (w) {
this.sums = [];
let sum = 0;
for (let v of w) {
sum += v;
this.sums.push(sum);
}
this.totalSum = sum;
};
/**
* @return {number}
*/
Solution.prototype.pickIndex = function () {
let target = Math.floor(Math.random() * this.totalSum);
let start = 0,
end = this.sums.length - 1;
while (start < end) {
let mid = Math.floor((start + end) / 2);
if (this.sums[mid] > target) end = mid;
else start = mid + 1;
}
return start;
};
/**
* Your Solution object will be instantiated and called as such:
* var obj = new Solution(w)
* var param_1 = obj.pickIndex()
*/
class Solution:
def __init__(self, w: List[int]):
self.sums = []
total_sum = 0
for weight in w:
total_sum += weight
self.sums.append(total_sum)
def pickIndex(self) -> int:
target = random.randint(1, self.sums[-1])
return bisect.bisect_left(self.sums, target)
# Your Solution object will be instantiated and called as such:
# obj = Solution(w)
# param_1 = obj.pickIndex()
class Solution {
public:
Solution(vector<int>& w) {
sums.push_back(w[0]);
for(int i=1; i<w.size() ; ++i){
sums.push_back(sums[i-1] + w[i]);
}
}
int pickIndex() {
int n = rand() % sums[sums.size()-1];
auto it = upper_bound(sums.begin(),sums.end(),n);
return it - sums.begin();
}
private:
vector<int> sums;
};
/**
* Your Solution object will be instantiated and called as such:
* Solution* obj = new Solution(w);
* int param_1 = obj->pickIndex();
*/
Time/Space Complexity:
- Time Complexity: O(n+Qlogn)
n
for initilization andlogn
forQ
queriespickIndex
- Space Complexity: O(n)
Explanation:
On initialization we build the accumulated sum, best seen on an example:
w = [3, 6, 4, 5]
=> sums = [3, 9, 13, 18]
which takes n
storage leading to linear space complexity. By doing this we enable every number to own the gap between itself and the previous element (e.g. 9
now owns the space between 4..9
, in total 6
numbers which is the original element in the input array for which we have accumulated sum of 9
).
On search we perform binary search for idx
randomly chosen from range 1..18
(random.nextInt(sums[sums.length-1]) + 1
). So if for example we get an index in range 4..9
the binary search will return either the exact index of element in array if idx == 9
or it will return insertion point (where the number would be inserted to maintain the sorted array, for this we're using built-in binary search) which again would be index 1
for range 4..9
. To find to which weight the range belongs we just index the initial weight array with the index we got from the binary search (for the range 4..9
weight 6
will be returned). The same goes for other ranges/gaps since for each index we will belong to one of the gaps which is owned by some element in sums/w
(initial array). Since we perform binary search, we have time complexity of O(logn)
. The built in binary search will return (-(insertionSort) - 1)
, to find the right in-boundary index we can use ~
which will do ~x = -x - 1
. So for the insertion point -2
which matches the second position in the array, we would have ~(-2) = -(-2) - 1 = 2 - 1 = 1
correctly mapping to the second number using index 1
.