CareerCup

Give a connected graph, no cycle,
Find the node where the average distance from all other nodes is the smallest

Input: an int array without sorting, then do a binary search on it, and find the number of digits that cannot be found by binary search.

<key = "a"; val = "3"; depend_list = null;/>; <key = "b"; val = "a * a";
Depend_list = {a};/>; <key = "c"; val = "a * b"; depend_list = {a,b};/>;
Lets output a map<key, value>,a -> 3, b -> 9, c -> 27
Follow-up questions: Make a statistic for all keys and corresponding map entries with the same value.

Given an array, find an x such that all numbers less than x are equal to themselves. Numbers greater than it are equal to x, and all numbers add up to a target
For example, [1 3 5 7 9] target=24, answer is8. Because when x=8, the array has only 9>8, so 1+3+5+7+8=24 is the closest to the target.

Give an N-length array with only 0 and 1 inside
Requires to find the minimum number of conversions needed to convert the array to 0 before all 1.
The conversion is to change a 0 to 1 or a 1 to 0,
And the number of 0 and 1 in the converted array can be arbitrary.
Just find the minimum conversion step

Imagine a horizontal corridor bounded by y = y1 and y = y2 lines on a 2D-plane. There are N sensors with centers (x, y) each of which operates in a range (r) on the plane . So, they cover some circular areas on the plane. See the figure below.

o            |
_____o___o__|____________ y = y1
		o	|OOO
__________oo|_____O______ y = y2
	O		|  O
_ _ _ _ _ _ | _ _ _ _ _ _
			|
	o O		|		o O
		O	|
			|
O			|
			|

The question is whether a path exist from x=-inf to x=+inf via corridor without being detected any radar.

Constraints:
1. You are free to move any direction only if you stay in the corridor.
2. You are free to go through corridor borders.
3. N sensors are given as list of (x, y, r) like [(1, 3, 2), (-1, -3, 4)]. x and y are signed integers. r is a unsigned integer.
4. y1 and y2 are integers.

My solution:
1. Create an intersection graph of N sensors by comparing them each other via Euclidean distance

(x1 - x2)^2 + (y1 - y2)^2 <= (r1 + r2)^2

2. If there is a path between y=y1 and y=y2 through intersected sensors, there do not exist any path from x=-inf to x=+inf. Otherwise, there do exist a path. So, I used BFS to search such a path.

Worst Case Complexity:
1. O(n^2)
2. O(V + E) = O(n + intersection_count)

Total: O(n^2)

My Best Case Optimization for Intersection Graph:
1. For each sensor, create two events for start and end of circles vertically. y = (y1 - r) and y = (y1 + r)
2. For each sensor, register those two events into an array.
3. Use a line sweep algorithm over 2, which is O(nlogn + intersection_count) and intersection_count may be n^2 at worst case.

I wonder if I should have had a better solution with the worst case < O(n^2).

Numbers with 4 are considered to be unlucky, floor numbers are skipped with numbers with 4; for a top level n, ask how many layers there are actually; for example n=20, that is 18 levels [Remove 4, 14]

I read this question here in careercup and saw no comments on it. So, I am answering it.
given a string find the sum of number of unique substring characters.
means if string is ACAX. the sub strings are A,C,A,X,AC,CA,AX,ACA,CAX,ACAX. in ACA, A occurs more than once so, its count is not considered, but C occurs once so its count is 1. Similarly in ACAX, A's count=0,C's count=1,X's count=1. In such a manner find sum of all counts for all substrings. The worst case time complexityis O(n). IF same substring occurs more than once then find its sum for that many times.

Input unsorted integer array represents a list of coins,
find the minimum amount of money that cannot be formed by these coins, each coin can only be used once
     E.g. {1,1} -> 3, {1,2,4} -> 8

Give me a list of int, find the length of the smallest cycle. For example, 1, 2, 1, 2, the length of the cycle is 2. Then 1, 2, 1, 2, 1 has a minimum length of 2. Then the length of 1, 2, 1, 2, 3 should be 5 because the entire list is not in repeat. Then the minimum length of 1, 2, 1, 2, 1, 1, 2 is 2.

Find the kth missing element in a sorted array. For example [2,3,5,7], k = 0: return 4, k = 1: return 6

expected time complexity logn

To an array and K, each element in the array can only move K positions to the left at most, no limit to the right, try to sort the array under the limit of K
a = [5, 2, 4, 3, 1], k = 2
Return [2, 3, 1, 4, 5]

Given a function rand32() (random number range 0-2**32-1) that randomly generates a 32-bit int, design a new random function:
Randn(n) generates a random distributed random number (0 - 2**n-1)
Rand3n(n) generates (0 - 3**n-1) the uniformly distributed random number

Topic: There is a set of coordinates. The original format of each coordinate is (1.3, 0.5). However, the comma and decimal point are gone. Only one string is left, allowing you to restore all possible combinations. For example, 123, possible (1, 23) (1, 2.3) (12, 3) (1.2, 3)

given period and threshold, Assume there is a endless streaming events, each event occurs at timestamp "x". The question want you to write an API that return true if number of the events are over the threshold within the period around timestamp "x"


Ex:
period = 3, threshold =2
getEvent(10) -> false
getEvent(12) -> false
getEvent(13) -> true [10,12,13]
getEvent(20) -> false


part one: event come in order
part two: event come without order

Given vector<int> nums, and pair<int, int> range. Find out how many continuous subsequences within this vector sum up the number within the range.

Input: [1, 2, 3], [3,6]
Output: (4)

because [1,2,3], [1,2], [2,3], [3]

Each have a (x,y) coordinate.
Write an API that group three Googler together for lunch if they are close enough. Otherwise, throw them in un-schedule pool.

Distance formula = sqrt((x1-x2) ^2 + (y1-y2) ^2)
Given an int range;

Range: 2
Input | Output of API Un-schedule pool
0,0 -> [] [[0,0]]
1,0 -> [] [[0,0], [1,0]]
3,0 -> [] [[0,0], [1,0], [3,0]]
1,1-> [[0,0], [1,0], [1,1]] [[3,0]]

Given an vector<int> nums and an int target, you can change any element of the vector to positive or negative. How many uniquely different vector sum up to target?
Input: [1,1,1], target = 2
[-1,1,1]
[1,-1,1]
[1,1,-1]

return (3)

find max path sum in DAG, weight can be negative

Implement a method to check if a n-ary tree is unival

class TreeNode {
int val;
List<TreeNode> children;
TreeNode(int val) {
this.val = val;
children = new ArrayList<>();
}
}

Given the length and width of a rectangle, how many ways can be used to go from the upper left corner to the upper right corner (each step can only go to the right, top right or bottom right):
-follow up 1: If three points in the rectangle must be visited, how many ways
-follow up 2: If you are given an point H, and the path must go down below the H point, how to do it

The thief has found himself a new place for his thievery again. There is only one entrance to this area, called the "root." Besides the root, each house has one and only one parent house. After a tour, the smart thief realized that "all houses in this place forms an n-nary tree". It will automatically contact the police if two directly-linked houses were broken into on the same night.
Determine the maximum amount of money the thief can rob tonight without alerting the police.
/**
* Definition for a n-ary tree node.
* public class TreeNode {
* int val;
* List<TreeNode> kids;
* }
*/
class Solution {
public int rob(TreeNode root) {

}
}

For two string a, b, their distance is defined as the number of positions at which the corresponding character are different.
Now you can swap characters at any two locations once, and ask how to swap to make the distance the smallest.

preorder Traversing a n-ary tree without using recurrsion

TreeNode<T> {
T val;
List<TreeNode> children;
}

Given two methods for the person class, one to find a dad and one to find a mother, using these two methods to achieve a method to determine whether the two people are related to blood, assuming a limited number of people.

Self-implemented data structures and methods, output all the heirs. To achieve birth (parent, name), dead (name), getAllSucession (). There is a king, you can use birth plus children, dead dead. The order of inheritance is the same as preorder except that this tree has multiple children.

Given an Input file of IPv4 addresses, filter and write them into Valid and Invalid IPs.

Input file = ["192.100.0.1, "10.0.0.1", "aa.bb.cc.dd", "10.0", "999.10.10.1"]
Valid = []
Invalid = []

A coffee machine has three buttons (S, M, L). After pressing each button, the coffee machine will flow out of a range of coffee [s1, s2], [m1, m2], [l1, l2], but the outflow of coffee The amount is random. There is a cup with a total capacity of c2. There is a tick c1 on the cup. If the amount of coffee is [c1, c2], it is considered full.
After asking for a series of button operations, can the coffee be filled in [c1, c2] but not overflow?

Input like method stack trace: "main, start", "foo, start", foo, end", "bar, start", "bar end", "main, end", "main2, start", "main2, end"
Output String, such as the following format, to indicate the order in which the method was called and the hierarchy

main
       Foo
       Bar
main2

public TreeNode{
int val = val;
TreeNode left, right;
public TreeNode(int val){
this.val = val;
}
given a tree and an API, delete a part of the nodes in the tree and return the forest formed after the node is deleted.
public List<TreeNode> deletenodes(TreeNode root, List<TreeNode> toDeletes){

}
example

1
 / \
 2  3
/  / \
4  5 6
If you delete the 2 and 5 nodes, you will need to return to the forest
[4,     1]
         \
          3
           \
           6