Develop Reference

the maximum n digit number possible in K steps

Can somebody help me with this problem?
Statement: - What is the maximum possible n digit number starting from 0 we can make in K steps
using only 2 operations:-
multiplying by 3 or incrementing by 2.
EXAMPLE :
N =2 K = 5;
-> (0->2->6->8->24->72) 72 IS THE ANSWER
N = 2 , K = 51 -> (0->2->6->8->10->30->32->96->98). 98 is the maximum we can get so need to check for rest of the moves.
My 2 state-recursive solution:-
public static void largestNDigitNumber(long[] highest, long maxValue, long k, long currentValue) {
if (highest[0] == (maxValue - 2)) return; //no need to do anything as we get 98 as highest.
if (k < 0) return; //checking for steps
if (highest[0] < currentValue && currentValue <= (maxValue - 2)) {
highest[0] = currentValue;
}
largestNDigitNumber(highest, maxValue, (k - 1), (currentValue * 3));
largestNDigitNumber(highest, maxValue, (k - 1), (currentValue + 2));
}
public static void main(String[] args) {
int n = 2;
long k = 51;
long maxValue = (long) Math.pow(10, n);
long[] highest = new long[1];
largestNDigitNumber(highest, maxValue, (k - 1), 2);
if (highest[0] < (long) Math.pow(10, (n - 1))) {
System.out.println("-1"); // if it is not possible to make n digit in given steps
} else System.out.println(highest[0]);
}
when "k" is small it is giving the correct answer but for bigger values of "k", it does not show any input. for n=2 and k = 51, it does not show anything.
please help me to improve this code

The question is equivalent to asking what is the largest base 3 number that is less than 10^n/2, and has digit sum plus length less than or equal to k+1. (The answer is then double the base 3 number).
For example, N=2 K=5. What's the largest base 3 number that's less than 50, with length plus digit sum less than or equal to 6. Answer: 1100 (36 decimal), so the answer to the original question is 36*2=72.
For N=2, K=51, the largest base-3 number that's less than 50 is 2001 (49 decimal) and has length sum plus digit sum = 7, which is way less than K+1.
Given this representation, it's easy to solve the problem in O(n) time (in fact, you can solve it using pencil and paper). The length d of the base-3 number is as large as possible such that 3^d < 10^n/2 and d<=K. Then fill in the digits of the number greedily from the most-significant first until you have digit sum K+1-d (or you run out of digits).
Equivalence
First note that without loss of generality you can assume you never have three +2 operations in a row, since that can be done more efficiently by inserting a single +2 operation to before the most recent *3 (or simply replacing it by +2 * 3 if there's no *3 operation). Suppose you have represented the current number as a doubled base-3 number. A +2 operation corresponds to adding 1 to the bottom digit (this never overflows into the next column thanks to the observation above). A *3 operation moves all the digits up one column, introducing a 0 as the bottom digit. Note that because the number is doubled, the +2 operation adds just 1 to the base-3 number!
From this, you can see that you can count the number of operations from observation of the doubled base-3 number. Because *3 introduces a new digit, and +2 increases the digit sum by 1, so the number of operations is equal to the number of digits plus 1, plus the digit sum.
As an example. Suppose you have the doubled base-3 number 2 * 2101, then this is equivalent to 2 * (1+3*3*(1+3*(1+1)))) = (2 + 3*3*(2+3*(2+2))).

I tried something like this. it seems to work fine.
getMaxNumber(2, 5) ==> 72
getMaxNumber(2, 51) ==> 98
private int getMaxNumber(int n, int k){
int N = 0;
for (int i = 0; i < n; i++) {
N = N * 10 + 9;
}
int[] result = new int[1];
helper(N, k, 0, 0, result);
return result[0];
}
private void helper(int N, int K, int n, int k, int[] result){
if(n > N) return;
if(k <= K){
result[0] = Math.max(result[0], n);
}
if(n > 0)
helper(N, K, n * 3, k + 1, result);
helper(N, K, n + 2, k + 1, result);
}

Keeping with the style of your original recursive method. I modified it a bit to produce a working solution:
public static long largestNDigitNumber(int n, long currentK, long maxK, long currentValue) {
if (currentK > maxK || n < 1 || maxK < 1) return 0;
if (currentValue >= Math.pow(10, n))
return 0;
long c1 = largestNDigitNumber(n, currentK + 1, maxK, currentValue * 3);
long c2 = largestNDigitNumber(n, currentK + 1, maxK, currentValue + 2);
if (c1 == 0 && c2 == 0)
return currentValue;
return c1 > c2 ? c1 : c2;
}
public static void main(String[] args) {
int n = 2;
long k = 51;
long largest = largestNDigitNumber(n, 0, k, 0);
System.out.println(largest); //98
}
This recursive method returns values here instead of using an array. Hence the check if one returned value is bigger than the other or they are both 0 before returning.

Both the +2 and *3 operations preserve odd/even parity, so starting from 0 we can only reach even numbers. We could start our search at the highest even number: 8, 98, 998, 9998 etc. and see what the shortest distance to 0 is.
If we are looking for the shortest distance, then there are less choices to make. If the current number is a multiple of 3 then there are two choices, either we divide by 3 or subtract 2. Otherwise the only choice is to subtract 2. I suspect that in the majority of cases, dividing by 3 is the better option, so that might be the first to try to keep the tree smaller.
If the minimum number of steps is less than K then as many divide by 3 operations as needed can be used to make the correct K
If the minimum number of steps is equal to K then the problem is solved.
If the minimum number of steps is more than K then you need to pick a lower starting number. Some even numbers will already have been covered as part of the initial calculation. You get those 'for free', provide you include a small amount of record keeping. You only need to examine large even numbers that were missed earlier due to a 'divide by 3' step.

Need explanation for one test case failed in Codility Peaks problem

A non-empty array A consisting of N integers is given.
A peak is an array element which is larger than its neighbors. More precisely, it is an index P such that 0 < P < N − 1, A[P − 1] < A[P] and A[P] > A[P + 1].
For example, the following array A:
A[0] = 1
A[1] = 2
A[2] = 3
A[3] = 4
A[4] = 3
A[5] = 4
A[6] = 1
A[7] = 2
A[8] = 3
A[9] = 4
A[10] = 6
A[11] = 2
has exactly three peaks: 3, 5, 10.
We want to divide this array into blocks containing the same number of elements. More precisely, we want to choose a number K that will yield the following blocks:
A[0], A1, ..., A[K − 1],
A[K], A[K + 1], ..., A[2K − 1],
...
A[N − K], A[N − K + 1], ..., A[N − 1].
What's more, every block should contain at least one peak. Notice that extreme elements of the blocks (for example A[K − 1] or A[K]) can also be peaks, but only if they have both neighbors (including one in an adjacent blocks).
The goal is to find the maximum number of blocks into which the array A can be divided.
Array A can be divided into blocks as follows:
one block (1, 2, 3, 4, 3, 4, 1, 2, 3, 4, 6, 2). This block contains three peaks.
two blocks (1, 2, 3, 4, 3, 4) and (1, 2, 3, 4, 6, 2). Every block has a peak.
three blocks (1, 2, 3, 4), (3, 4, 1, 2), (3, 4, 6, 2). Every block has a peak. Notice in particular that the first block (1, 2, 3, 4) has a peak at A[3], because A[2] < A[3] > A[4], even though A[4] is in the adjacent block.
However, array A cannot be divided into four blocks, (1, 2, 3), (4, 3, 4), (1, 2, 3) and (4, 6, 2), because the (1, 2, 3) blocks do not contain a peak. Notice in particular that the (4, 3, 4) block contains two peaks: A[3] and A[5].
The maximum number of blocks that array A can be divided into is three.
Write a function:
class Solution { public int solution(int[] A); }
that, given a non-empty array A consisting of N integers, returns the maximum number of blocks into which A can be divided.
If A cannot be divided into some number of blocks, the function should return 0.
For example, given:
A[0] = 1
A[1] = 2
A[2] = 3
A[3] = 4
A[4] = 3
A[5] = 4
A[6] = 1
A[7] = 2
A[8] = 3
A[9] = 4
A[10] = 6
A[11] = 2
the function should return 3, as explained above.
Write an efficient algorithm for the following assumptions:
N is an integer within the range [1..100,000];
each element of array A is an integer within the range [0..1,000,000,000].
My Understanding of the problem :
Each sub array should contain at least one peak
An element which forms a peak can be in an Adjacent sub array.
Return max possible sub arrays
My Question
Consider Main Array : [0,1,0,1,0]
Possible sub arrays as per understanding : [0,1] [0,1,0]
Each subarray has a peak.
Subarray 1 [0,1] has peak element shared with adjacent array [0,1,0].
Subarray 2 [0,1,0] contains peak 0<1>0.
So max possible sub arrays are 2 but a test case in Codility returns max possible sub arrays as 1.
Below is my code
// you can also use imports, for example:
import java.util.*;
// you can write to stdout for debugging purposes, e.g.
// System.out.println("this is a debug message");
class Solution {
public int solution(int[] A) {
// write your code in Java SE 8
int count=0,size=A.length;
if(size<2)
return 0;
System.out.println(Arrays.toString(A));
for(int i=1;i<size-1;i++){
if(A[i-1]<A[i] && A[i]>A[i+1]){
count++;
i++;
}
}
return count;
}
}
Test case which failed in Codility click here
I believe there is a gap in my understanding. Any help would be helpful :)

https://app.codility.com/demo/results/training5KP2PK-P4M/
https://github.com/niall-oc/things/blob/master/codility/peaks.py
Breaking an array into even parts is another way of factorizing the length of the array.
Array of length 12
[0,1,0,1,0,0,0,1,0,0,1,0,]
Factors of 12.
The square root of 12 is 3.464... so start with 3 and irterate down to 1 and divide each number into 12. This gives you a set of numbers {1, 2, 3, 4, 6, 12}.
Process
Because of how a peak is defined you cannot have 12 peaks in this array so remove 12 from the set. Starting d as the largest number divide the array int d parts. And check each part has a peak, if so then d is the maximum number of equal parts to all contain at least one peak. If not so then iterate to the next largest divisor and try that until you find a solution.

First of all you need to be congratulated to write a concise program to calculate the number of peaks. But the question is not to count the peaks.
It is to find the number of equal sized array and each having at least one peak. And a peak cannot be the first or last element as stated in the problem 0 < P < N − 1
Now quoting your question:
Consider Main Array : [0,1,0,1,0]
Possible sub arrays as per understanding : [0,1] [0,1,0] Each subarray has a peak. Subarray 1 [0,1] has peak element shared with adjacent array [0,1,0]. Subarray 2 [0,1,0] contains peak 0<1>0.
So max possible sub arrays are 2 but a test case in Codility returns max possible sub arrays as 1.
I see below issues:
your sub array sizes are not equal
array [0,1] does not have a peak
So, the array cannot be divided in equal parts each having a peak and hence only one array remains [0,1,0,1,0].

How does this shuffling with Math rand work?

I saw this code to shuffle a list:
public static void shuffle(List<Integer> numbers) {
if(numbers == null || numbers.isEmpty()) return;
for(int i = 0; i < numbers.size(); ++i) {
int index = (int) (i + Math.random()*(numbers.size() - i));
swap(numbers, i, index);
}
}
The code seem to work but I don't understand this snippet:
int index = (int) (i + Math.random()*(numbers.size() - i));
Basically it is i + R*(n-i) but how does this ensure that: i) we won't get an out of bounds index or ii) I won't be changing the same element's i.e. index == i and the shuffle would not be that random?

Math.random() returns a uniform random number in the interval [0, 1), and numbers.size() - i, ideally, scales that number to the interval [0, numbers.size() - i). For example, if i is 2 and the size of the list is 5, a random number in the interval [0, 3) is chosen this way, in the ideal case. Finally, i is added to the number and the (int) cast discards the number's fractional part. Thus, in this example, a random integer in [2, 5) (that is, either 2, 3, or 4) is generated at random, so that at each iteration, the number at index X swaps with itself or a number that follows it.
However, there is an important subtlety here. Due to the nature of floating-point numbers and rounding error when scaling the number, in extremely rare cases the output of Math.random()*(numbers.size() - i) might be equal to numbers.size() - i, even if Math.random() outputs a number that excludes 1. rounding error can cause the idiom Math.random()*(numbers.size() - i) to bias some results over others. For example, this happens whenever 2^53 is not divisible by numbers.size() - i, since Math.random() uses java.util.Random under the hood, and its algorithm generates numbers with 53 bits of precision. Because of this, Math.random() is not the best way to write this code, and the code could have used a method specially made for generating random integers instead (such as the nextInt method of java.util.Random). See also this question and this question.
EDIT: As it turns out, the Math.random() * integer idiom does not produce the issue that it may return integer, at least when integer is any positive int and the round-to-nearest rounding mode is used as in Java. See this question.

Math.random() always returns a floating-point number between 0 (inclusive) and 1 (exclusive). So when you do Math.random()*(numbers.size() - i), the result will always be between 0 (inclusive) and n-i (exclusive).
Then you add i to it in i + Math.random()*(numbers.size() - i).
Now the result, as you can see, will be between i (inclusive) and n (exclusive).
After that, you are casting it to an int. When you cast a double to an int, you truncate it, so now the value of index will somewhere from ``iton - 1``` (inclusive for both).
Therefore, you will not have an ArrayIndexOutOfBoundsException, since it will always be at least 1 less than the size of the array.
However, the value of index could be equal to i, so yes, you are right in that a number could be swapped with itself and stay right there. That's perfectly fine.

You have a list of 1 to 50 ints.
So get a random value from 0 to 49 inclusive to index it.
say it is 30.
Get item at index 30.
Now replace item at index 30 with item at index 49.
Next time generate a number between 0 and 48 inclusive. 49 will never be reached and the number that was there occupies the slot of the last number used.
Continue this process until you've exhausted the list.
Note: that the expression (int)(Math.random() * n) will generate a random number between 0 and n-1 inclusive because Math.random generates a number between 0 and 1 exclusive.

Instead of using such a custom method, I recommend you use OOTB Collections.shuffle. Check this to understand the logic implemented for Collections.shuffle.
Analysis of your code:
Math.random() returns a double value with a positive sign, greater than or equal to 0.0 and less than 1.0.
Now, let's assume numbers.size() = 5 and dry run the for loop:
When i = 0, index = (int) (0 + Math.random()*(5 - 0)) = (int) (0 + 4.x) = 4
When i = 1, index = (int) (1 + Math.random()*(5 - 1)) = (int) (1 + 3.x) = 4
When i = 2, index = (int) (2 + Math.random()*(5 - 2)) = (int) (2 + 2.x) = 4
When i = 3, index = (int) (3 + Math.random()*(5 - 3)) = (int) (3 + 1.x) = 4
When i = 4, index = (int) (4 + Math.random()*(5 - 4)) = (int) (4 + 0.x) = 4
As you can see, the value of index will remain 4 in each iteration when numbers.size() = 5.
Your queries:
how does this ensure that: i) we won't get an out of bounds index
As already explained above using the dry run, it will never go out of bounds.
or ii) I won't be changing the same element's i.e. index == i and the
shuffle would not be that random?
swap(numbers, i, index); is swapping the element at index, i with the element at index, 4 each time when numbers.size() = 5. This is illustrated with the following example:
Let's say numbers = [1, 2, 3, 4, 5]
When i = 0, numbers will become [5, 2, 3, 4, 1]
When i = 1, numbers will become [5, 1, 3, 4, 2]
When i = 2, numbers will become [5, 1, 2, 4, 3]
When i = 3, numbers will become [5, 1, 2, 3, 4]
When i = 4, numbers will become [5, 1, 2, 3, 4]

int index = (int) (i + Math.random()*(numbers.size() - i)); - it is important to note that Math.random() will generate a number which belongs to <0;1). So it will never exceed the boundry as exclusive max will be: i + 1*(number.size() -i) = number.size
This point is valid, it can happen.

How to get the kth term in N Combination R [closed]

Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 4 years ago.
Improve this question
How do I get the kth combination inNCR. without iterating through all possible outcomes. e.g. say I have 3C2 for 3 positions and 2identical-items. I am aware it's [011],[101] and [110]. how do I get e.g. the 2nd term(k=1) which is [101] using a method?
constraints(R < N k >= 0 and k < P where P = NCR).
NB:[101] is the 2nd term(in ascending/lexicographical order) because 011 = 3,101 = 5 ,110 = 6
in decimal. so basically the goal is to get what number k in NCR is,
because every kth output from NCR can be represented as a number.

Yes, you are correct when you say:
because every kth output from NCR can be represented as a number.
There is a bijection from the set of integers 1 to # of combs/perms to the entire set of combs/perms. Finding the specific index of a particular comb/perm is sometimes referred to as getting the rank. According to the example that you have in your question, these are ordinary permutations. Moreover when you mention ascending order, you are referring to the lexicographical order.
It is a straightforward exercise in counting to obtain the nth ordinary permutation of a given set. We first need to obtain the total number of permutations using the well established formula:
P(n, r) = n! / (n - r)!
This next part is the key observation that allows us to quickly obtain each element of our target permutation.
If we look at all permutations of our set of n choose r, there will be n groups that are only different by a permutation of the n elements.
For example, if we look at the first two group of the permutations of [0 1 2 3] choose 3, we have:
[,0] [,1] [,2]
[0,] 0 1 2
[1,] 0 1 3
[2,] 0 2 1
[3,] 0 2 3
[4,] 0 3 1
[5,] 0 3 2
[6,] 1 0 2
[7,] 1 0 3
[8,] 1 2 0
[9,] 1 2 3
[10,] 1 3 0
[11,] 1 3 2
Note that the last permutations are simply the first 6 permutations of the set [1 0 2 3].. that is, 0 is mapped to 1, 1 is mapped to 0, and the final 2 elements are mapped to themselves.
This pattern continues as we move to the right only instead of n identical groups, we will get n - 1 similar groups for the second column, n -2 for the third, and so on.
So to determine the first element of our permutation, we need to determine the 1st group. We do that by simply dividing the number of permutations by n. For our example above of permutations of 4 choose 3, if we were looking for the 15th permutation, we have the following for the first element:
Possible indices : [0 1 2 3]
P(4, 3) = 24
24 / 4 = 6 (elements per group)
15 / 6 = 2 (integer division) 2 means the 3rd element here (base zero)
Now that we have used the 3rd element, we need to remove it from our array of possible indices. How do we get the next element?
Easy, we get our next subindex by subtracting the product of the group we just found and the elements per group from our original index.
Possible indices : [0 1 3]
Next index is 15 - 6 * 2 = 3
Now, we just repeat this until we have filled all entries:
Possible indices : [0 1 3]
Second element
6 / 3 = 2 (elements per group)
3 / 2 = 1
Next index is 3 - 3 * 1 = 0
Possible indices : [0 3]
Third element
2 / 2 = 1
0 / 1 = 0
So our 15th element is : [2 1 0]
Here is a C++ implementation that should be pretty easy to translate to Java:
double NumPermsNoRep(int n, int k) {
double result = 1;
double i, m = n - k;
for (i = n; i > m; --i)
result *= i;
return result;
}
std::vector<int> nthPermutation(int n, int r, double myIndex) {
int j = 0, n1 = n;
double temp, index1 = myIndex;
std::vector<int> res(r);
temp = NumPermsNoRep(n, r);
std::vector<int> indexVec(n);
std::iota(indexVec.begin(), indexVec.end(), 0);
for (int k = 0; k < r; ++k, --n1) {
temp /= n1;
j = (int) std::trunc(index1 / temp);
res[k] = indexVec[j];
index1 -= (temp * (double) j);
indexVec.erase(indexVec.begin() + j);
}
}
These concepts extends to other types of combinatorial problems, such as finding the nth combination, or permutation with repetition, etc.

The time complexity is O(kn), space is O(n)
public static void main(String[] args) {
//n = 4, r = 2, k = 3
int[] ret1 = getKthPermutation(4, 2, 3);
//ret1 is [1,0,0,1]
//n = 3, r = 2, k = 1
int[] ret2 = getKthPermutation(3, 2, 1);
//ret2 is [1,0,1]
}
static int[] getKthPermutation(int n, int r, int k) {
int[] array = new int[n];
setLastN(array, r, 1);
int lastIndex = n - 1;
for(int count = 0; count < k; count++) {
int indexOfLastOne = findIndexOfLast(array, lastIndex, 1);
int indexOfLastZero = findIndexOfLast(array, indexOfLastOne, 0);
array[indexOfLastOne] = 0;
array[indexOfLastZero] = 1;
//shortcut: swap the part after indexOfLastZero to keep them sorted
int h = indexOfLastZero + 1;
int e = lastIndex;
while(h < e) {
int temp = array[h];
array[h] = array[e];
array[e] = temp;
h++;
e--;
}
}
return array;
}
//starting from `from`, and traveling the array forward, find the first `value` and return its index.
static int findIndexOfLast(int[] array, int from, int value) {
for(int i = from; i > -1; i--)
if(array[i] == value) return i;
return -1;
}
//set the last n elements of an array to `value`
static void setLastN(int[] array, int n, int value){
for(int i = 0, l = array.length - 1; i < n; i++)
array[l - i] = value;
}
This is an adaption of the very typical "find the kth permation" algorithm.
I will try to explain the general idea (yours is a special case as there are only two types of elements: 0 and 1).
Lets say I have [2,1,6,4,7,5]. What is the next smallest permutation that is bigger than the current one? Why do I concern the next smallest permutation bigger than current one? Because if you start with the smallest permutation [1,2,4,5,6,7] and you repeat the action (find the smallest bigger than current) k times, you will find k+1 th smallest permutation.
Now, since the one I am looking for needs to be bigger than current one, I need to increment the current one. To keep the incrementation as small as possible, I am going to try to modify 5 (last one). Now, I cannot just change 5 to a random value, I can only swap it with some digit before it.
If I swap 5 with a bigger number before it, say 7, then I will get [2,1,6,4,5,7], which is smaller than current one. Now obviously I need to swap 5 with some smaller digit before it, but which one? If I swap 5 with 2, I get [5,1,6,4,7,2], this increment is too big. I need to swap 5 with a "lower digit" to keep the increment as small as possible. Thats leads us to find the first(lowest) digit (from right to left) that is smaller than 5. In this case I would need to swap 5 with 4 and get [2,1,6,5,7,4]. This way, I can make the impact of "swap" small. Now the prefix is decided [2,1,6,5. There is no smaller prefix. We need to deal with suffix 7,4]. Clearly, if we sort the suffix and make it 4,7], then we are done.
In our case, there are two differences:
1. we need to swap the last 1, because you cannot make the permutation bigger by swapping the a zero with any digit before it.
2. we can always sort the suffix using a shortcut as shown in the code. I will leave it to you:)

public static String lexicographicPermutation(String str, long n) {
final long[] factorials = { 1, 2, 6, 24, 120, 720, 5040, 40320, 362880, 3628800, 39916800, 479001600 };
n--;
char[] arr = str.toCharArray();
for (int i = 0; i < arr.length - 1; i++) {
long fact = factorials[arr.length - i - 2];
long p = i + n / fact;
n %= fact;
for (int j = i + 1; j <= p; j++)
swap(arr, i, j);
}
return new String(arr);
}
private static void swap(char[] arr, int i, int j) {
char tmp = arr[i];
arr[i] = arr[j];
arr[j] = tmp;
}
You can replace STR with required string. In the given example, 1st permutation is "abcdefghijklm" (this is a string with 13 chars), 13!st permutation is reverse string "mlkjihgfedcba" and 100st permutation is "abcfklgmeihjd".
To realise this soulution just google Factorial number system. This is a key to solve this problem. This is a Project Euler: Problem 24.
Demo:
for(int i = 1; i <= 6; i++)
System.out.println(lexicographicPermutation("110", i));
1 - 110
2 - 101
3 - 110
4 - 101
5 - 011
6 - 011
for(int i = 1; i <= 6; i++)
System.out.println(lexicographicPermutation("abc", i));
1 - abc
2 - acb
3 - bac
4 - bca
5 - cab
6 - cba

Expected number of maxima

I have is algorithm, which takes an array as an argument, and returns its maximum value.
find_max(as) :=
max = as[0]
for i = 1 ... len(as) {
if max < as[i] then max = as[i]
}
return max
My question is: given that the array is initially in a (uniformly) random permutation and that all its elements are distinct, what's the expected number of times the max variable is updated (ignoring the initial assignment).
For example, if as = [1, 3, 2], then the number of updates to max would be 1 (when reading the value 3).

Assume the original array contains the values 1, 2, ..., N.
Let X_i, i = 1..N be random variables that take the value 1 if i is, at some point during the algorithm, the maximum value.
Then the number of maximums the algorithm takes is the random variable: M = X_1 + X_2 + ... + X_N.
The average is (by definition) E(M) = E(X_1 + X_2 + ... + X_N). Using linearity of expectation, this is E(X_1) + E(X_2) + .. + E(X_N), which is prob(1 appears as a max) + prob(2 appears as a max) + ... + prob(N appears as a max) (since each X_i takes the value 0 or 1).
When does i appear as a maximum? It's when it appears first in the array amongst the i, i+1, i+2, ..., N. The probability of this is 1/(N-i+1) (since each of those numbers are equally likely to be first).
So... prob(i appears as a max) = 1/(N-i+1), and the overall expectation is 1/N + 1/(N-1) + ..+ 1/3 + 1/2 + 1/1
This is Harmonic(N) which is approximated closely by ln(N) + emc where emc ~= 0.5772156649, the Euler-Mascheroni constant.
Since in the problem you don't count the initial setting of the maximum to the first value as a step, the actual answer is Harmonic(N) - 1, or approximately ln(N) - 0.4227843351.
A quick check for some simple cases:
N=1, only one permutation, and no maximum updates. Harmonic(1) - 1 = 0.
N=2, permutations are [1, 2] and [2, 1]. The first updates the maximum once, the second zero times, so the average is 1/2. Harmonic(2) - 1 = 1/2.
N=3, permutations are [1, 2, 3], [1, 3, 2], [2, 1, 3], [2, 3, 1], [3, 1, 2], [3, 2, 1]. Maximum updates are 2, 1, 1, 1, 0, 0 respectively. Average is (2+1+1+1)/6 = 5/6. Harmonic(3) - 1 = 1/2 + 1/3 = 5/6.
So the theoretical answer looks good!

Empirical Solution
A simulation of many different array sizes with multiple trials each can be performed and analyzed:
#include <iostream>
#include <fstream>
#include <cstdlib>
#define UPTO 10000
#define TRIALS 100
using namespace std;
int arr[UPTO];
int main(void){
ofstream outfile ("tabsep.txt");
for(int i = 1; i < UPTO; i++){
int sum = 0;
for(int iter = 0; iter < TRIALS; iter++){
for(int j = 0; j < i; j++){
arr[j] = rand();
}
int max = arr[0];
int times_changed = 0;
for(int j = 0; j < i; j++){
if (arr[j] > max){
max = arr[j];
times_changed++;
}
}
sum += times_changed;
}
int avg = sum/TRIALS;
outfile << i << "\t" << avg << "\n";
cout << "\r" << i;
}
outfile.close();
cout << endl;
return 0;
}
When I graphed these results, the complexity appeared to be logarithmic:
I think it's safe to conclude that the time complexity is O(log n).
Theoretical solution:
Assume that the numbers are in the range 0...n
You have a tentative maximum m
The next maximum will be a random number in the range m+1...n, which averages out to be (m+n)/2
This means that each time you find a new maximum, you are dividing the range of possible maximums by 2
Repeated division is equivalent to a logarithm
Therefore the number of times a new maximum is found is O(log n)

Worst case scenario (which is often what is sought) is O(n). If the list is sorted in reverse order every single one will result in an assignment.
HOWEVER, if your assignment is the most expensive operation why don't you just store it's index and only ever copy once, if at all? In that case, you will have exactly 1 assignment and n-1 comparisons.