
Data Structure
Networking
RDBMS
Operating System
Java
MS Excel
iOS
HTML
CSS
Android
Python
C Programming
C++
C#
MongoDB
MySQL
Javascript
PHP
- Selected Reading
- UPSC IAS Exams Notes
- Developer's Best Practices
- Questions and Answers
- Effective Resume Writing
- HR Interview Questions
- Computer Glossary
- Who is Who
Find KPR Sum for All Queries in Python
Suppose we have a list of numbers nums. We also have a list of queries where queries[i] contains three elements [k, p, r], for each query we shall have to find kpr_sum. The formula for kpr_sum is like below.
$$\mathrm{{???}\_{???} =\sum_{\substack{?=?}}^{?−1}\sum_{\substack{?=?+1}}^{?}(? ⊕(?[?]⊕?[?]))}$$
If the sum is too large, then return sum modulo 10^9+7.
So, if the input is like nums = [1,2,3] queries = [[1,1,3],[2,1,3]], then the output will be [5, 4] because for the first element it is (1 XOR (1 XOR 2)) + (1 XOR (1 XOR 3)) + (1 XOR (2 XOR 3)) = 5, similarly for second query, it is 4.
To solve this, we will follow these steps −
- m := 10^9 + 7
- N := size of nums
- q_cnt := size of queries
- C := a new list
- res := a new list
- for i in range 0 to 19, do
- R := an array with single element 0
- t := 0
- for each x in nums, do
- t := t + (x after shifting i times to the right) AND 1
- insert t at the end of R
- insert R at the end of C
- for j in range 0 to q_cnt, do
- (K, P, R) := queries[j]
- d := R - P + 1
- t := 0
- for i in range 0 to 19, do
- n1 := C[i, R] - C[i, P-1]
- n0 := d - n1
- if (K after shifting i times to the right) AND 1 is non-zero, then
- x := quotient of (n1 *(n1 - 1) + n0 *(n0 - 1))/2
- otherwise,
- x := n1 * n0
- t :=(t +(x after shifting i times to the left)) mod m
- insert t at the end of res
- return res
Example
Let us see the following implementation to get better understanding −
def solve(nums, queries): m = 10**9 + 7 N = len(nums) q_cnt = len(queries) C = [] res = [] for i in range(20): R = [0] t = 0 for x in nums: t += (x >> i) & 1 R.append(t) C.append(R) for j in range(q_cnt): K, P, R = queries[j] d = R - P + 1 t = 0 for i in range(20): n1 = C[i][R] - C[i][P-1] n0 = d - n1 if (K >> i) & 1: x = (n1 * (n1 - 1) + n0 * (n0 - 1)) >> 1 else: x = n1 * n0 t = (t + (x << i)) % m res.append(t) return res nums = [1,2,3] queries = [[1,1,3],[2,1,3]] print(solve(nums, queries))
Input
[1,2,3], [[1,1,3],[2,1,3]]
Output
[5, 4]
Advertisements