↳

my solution was still O(n) - 3n to be exact. i "iterate" over the entire matrix 3 times - once to check for duplicates in the rows, once to check for duplicates in the columns, and once to check for duplicates within sub matrices. During each of those iterations I visit each element of the matrix exactly once and just check whether it's already in a dictionary of seen numbers which would be a constant time operation, so this is in total would be O(n). I can't think of how to check all those in just one iteration , if that's what you were saying you could do. (at least without using O(n) extra space.. mine only uses constant extra space) i would appreciate if you could elaborate Minder

↳

I basically just checked for duplicate numbers in each row, column, and then each 3x3 inner matrix. each iteration, i reset a dictionary of "seen" numbers. "." characters wouldn't matter, if i saw a number 1-9 i would check if it was already in the dictionary, if it was return false. return true at the very end if never encountered repeat numbers in rows, columns, or inner squares. when I explained this approach to the interviewer and he seemed satisfied with it although it was not easy to tell what the hell he was thinking or saying. i was working in a shared text file (i think it was codepen?), and even though the interviewer could have typed too, he didn't (even when i had to ask him repeatedly to spell a word i couldn't make out).. Minder

↳

1) Subtract each number from summation formula = N * (N+1) / 2 2) Hash table , zero out all entries when insert number -> 1 after all insert look for 0 entry 3) XOR all elems -> X XOR all # 1 to N -> Y XOR of X and Y -> missing # 15 ^ 12 ^ 15 = 12 Minder

↳

Stupid Math this is called stalking not algorithm

↳

With 360 degrees in a circle and 12 hours throughout each cycle the distance between each number is 30 degrees. If the minute hand is pointing at the 3 and is fifteen minutes, fifteen minutes is 1/4 of the total amount of minutes on the clock, this puts the hour hand 30/4 degrees passed the 3 at 7.25 degrees. They'll enjoy you thinking analytically. Minder

↳

The answer should be a right angle.

↳

can't u just parse char by char and use a bunch of shift operators and since ur just using charAt(i) to pick up a character u can use the same i and just raise it to the power of 10 and just or it with some bytes Minder

↳

public static int parseInt(String stringToConvert) { int i =0,num=0; int isNeg = 1; int length = stringToConvert.length(); if(stringToConvert.charAt(0) =='-') { isNeg=-1; i=1; } while(i Minder

↳

It's actually a pretty simple dynamic programming problem you can solve using constant space, do it iteratively rather than recursively. Check it out of LeetCode. Minder

↳

This can be calculated with a simple recursive Fibonacci

↳

int ans[] = new int[1]; //O(n) public int efficientDia(TreeNode root) { if(root == null) return 0; int left = efficientDia(root.left); int right = efficientDia(root.right); ans[0] = Math.max(ans[0], 1 + left+ right); return 1+ Math.max(left, right); } //O(n^2) public int getDiameter(TreeNode root) { if(root == null) return 0; int leftHeight = getHeight(root.left); int rightHeight = getHeight(root.right); if(ans[0] < 1 + leftHeight + rightHeight) { ans[0] = 1 + leftHeight + rightHeight; } return Math.max(getDiameter(root.left), getDiameter(root.right)); } private int getHeight(TreeNode root) { // TODO Auto-generated method stub if(root == null) return 0; return Math.max(getHeight(root.left), getHeight(root.right))+1; } Minder

↳

Indians at all companies always ask tree questions, it makes them giggle inside. I know, because I'm half indian and have interviewed people... tee hee hee Minder

↳

Yes they ask you which takes up less memory and there are multiple answers due to the number of elements (as you've stated) as well as the type of the data being stored. A hash table would be better if there aren't many elements since you typically have more slots than you have actual elements so you aren't resizing on every insert. Similarly if you're storing larger object say of 64 bytes, then the extra pointer data stored in a basic tree doesn't amount to much in comparison. However if you're storing something like a character or even a 32-bit integer, a binary tree will take more space since each node (in a basic implementation) needs to store (assuming 32-bit pointer) 64-bits with of pointers per element. Point is the question is stupid without further information since the answer depends on information they don't give you. But they do expect you to use a binary tree. Minder

↳

If minimizing memory footprint is the only requirement, I would use nothing, since nothing doesn't take up any memory and satisfies the functional requirement. Are you asking me which of these takes up less memory? That depends on the number of elements in the storage. But usually, a hash-table has less memory overhead Trees have overhead associated with every node. Minder

↳

public void printNumbers () { for (int i = 1; i <= 100; ++i) { if (i % 15==0) { System.out.println("FooBizz"); } else if (i % 3 == 0) { System.out.println("Foo"); } else if (i % 5 == 0) { System.out.println("Bizz"); } else { System.out.println(i); } } } Minder

↳

In these sorts of interviews you really need to drill down and understand what the interviewer is looking for. A good way to simulate a real interview experience is to do a mock with one of the Envoy Mobile Software Engineer experts on Prepfully, rated super strongly on TrustPilot... prepfully.com/practice-interviews Minder