Two players take turns choosing one of the outer coins. At the end we calculate the difference
between the score two players get, given that they play optimally.
The greedy strategy of getting the max. value of coin often does not lead to the best results in my case.
Now I developed an algorithm:
Sample:{9,1,15,22,4,8}
We calculate the sum of coins in even index and that of coins in odd index.
Compare the two sum, (9+15+4)<(1+22+8) so sum of odd is greater. We then pick the coin with odd index, in our sample that would be 8.
the opponent, who plays optimally, will try to pick the greater coin, e.g. 9.
There is always a coin at odd index after the opponent finished, so we keep picking the coins
at odd index, that would be 1.
looping the above steps we will get a difference of (8+1+22) - (9+15+4) = 3.
6.vice versa if sum of even is greater in step 2.
I have compared the results generated by my algorithm with a 2nd algorithm similar to below one: https://www.geeksforgeeks.org/optimal-strategy-for-a-game-set-2/?ref=rp
And the results were congruent, until my test generated a random long array:
[6, 14, 6, 8, 6, 3, 14, 5, 18, 6, 19, 17, 10, 11, 14, 16, 15, 18, 7, 8, 6, 9, 0, 15, 7, 4, 19, 9, 5, 2, 0, 18, 2, 8, 19, 14, 4, 8, 11, 2, 6, 16, 16, 13, 10, 19, 6, 17, 13, 13, 15, 3, 18, 2, 14, 13, 3, 4, 2, 13, 17, 14, 3, 4, 14, 1, 15, 10, 2, 19, 2, 6, 16, 7, 16, 14, 7, 0, 9, 4, 9, 6, 15, 9, 3, 15, 11, 19, 7, 3, 18, 14, 11, 10, 2, 3, 7, 3, 18, 7, 7, 14, 6, 4, 6, 12, 4, 19, 15, 19, 17, 3, 3, 1, 9, 19, 12, 6, 7, 1, 6, 6, 19, 7, 15, 1, 1, 6]
My algorithm generated 26 as the result, while the 2nd algorithm generated 36.
Mine is nothing about dynamic programming and it requires less memory, whereas i also implemented the 2nd one with memoization.
This is confusing since mine is correct with most of the array cases until this one.
Any help would be appreciated!
If the array is of even length, your algorithm tries to produce a guaranteed win. You can prove that quite easily. But it doesn't necessarily produce the optimal win. In particular it won't find strategies where you want some coins that are on even indexes and others on odd indexes.
The following short example illustrates the point.
[10, 1, 1, 20, 1, 1]
Your algorithm will look at evens vs odds, realize that 10+1+1 < 1+20+1 and take the last element first. Guaranteeing a win by 10.
But you want both the 10 and the 20. Therefore the optimal strategy is to take the 10 leaving 1, 1, 20, 1, 1, whichever side the other person takes you take the other to get to 1, 20, 1, and then whichever side the other takes you take the middle. Resulting in you getting 10, 1, 20 and the other person getting 1, 1, 1. Guaranteeing a win by 28.
Related
Assume I have two arrays, both of them are sorted, for example:
A: [1, 4, 5, 8, 10, 24]
B: [3, 6, 9, 29, 50, 65]
And then I merge these two array into one array and keep original relative order of both two array
C: [1, 4, 3, 5, 6, 9, 8, 29, 10, 24, 50, 65]
Is there any way to split C into two sorted array in O(n) time?
note: not necessarily into the original A and B
Greedily assign your integers to list 1 if they can go there. If they can't, assign them to list 2.
Here's some Ruby code to play around with this idea. It randomly splits the integers from 0 to n-1 into two sorted lists, then randomly merges them, then applies the greedy approach.
def f(n)
split1 = []
split2 = []
0.upto(n-1) do |i|
if rand < 0.5
split1.append(i)
else
split2.append(i)
end
end
puts "input 1: #{split1.to_s}"
puts "input 2: #{split2.to_s}"
merged = []
split1.reverse!
split2.reverse!
while split1.length > 0 && split2.length > 0
if rand < 0.5
merged.append(split1.pop)
else
merged.append(split2.pop)
end
end
merged += split1.reverse
merged += split2.reverse
puts "merged: #{merged.to_s}"
merged.reverse!
greedy1 = [merged.pop]
greedy2 = []
while merged.length > 0
if merged[-1] >= greedy1[-1]
greedy1.append(merged.pop)
else
greedy2.append(merged.pop)
end
end
puts "greedy1: #{greedy1.to_s}"
puts "greedy2: #{greedy2.to_s}"
end
Here's sample output:
> f(20)
input 1: [2, 3, 4, 5, 8, 9, 10, 18, 19]
input 2: [0, 1, 6, 7, 11, 12, 13, 14, 15, 16, 17]
merged: [2, 0, 1, 6, 3, 4, 5, 8, 9, 7, 10, 11, 18, 12, 13, 19, 14, 15, 16, 17]
greedy1: [2, 6, 8, 9, 10, 11, 18, 19]
greedy2: [0, 1, 3, 4, 5, 7, 12, 13, 14, 15, 16, 17]
> f(20)
input 1: [1, 3, 5, 6, 8, 9, 10, 11, 13, 15]
input 2: [0, 2, 4, 7, 12, 14, 16, 17, 18, 19]
merged: [0, 2, 4, 7, 12, 14, 16, 1, 3, 5, 6, 8, 17, 9, 18, 10, 19, 11, 13, 15]
greedy1: [0, 2, 4, 7, 12, 14, 16, 17, 18, 19]
greedy2: [1, 3, 5, 6, 8, 9, 10, 11, 13, 15]
> f(20)
input 1: [0, 1, 2, 6, 7, 9, 11, 14, 15, 18]
input 2: [3, 4, 5, 8, 10, 12, 13, 16, 17, 19]
merged: [3, 4, 5, 8, 10, 12, 0, 13, 16, 17, 1, 19, 2, 6, 7, 9, 11, 14, 15, 18]
greedy1: [3, 4, 5, 8, 10, 12, 13, 16, 17, 19]
greedy2: [0, 1, 2, 6, 7, 9, 11, 14, 15, 18]
Let's take your example.
[1, 4, 3, 5, 6, 9, 8, 29, 10, 24, 50, 65]
In time O(n) you can work out the minimum of the tail.
[1, 3, 3, 5, 6, 8, 8, 10, 10, 24, 50, 65]
And now the one stream is all cases where it is the minimum, and the other is the cases where it isn't.
[1, 3, 5, 6, 8, 10, 24, 50, 65]
[ 4, 9, 29, ]
This is all doable in time O(n).
We can go further and now split into 3 streams based on which values in the first stream could have gone in the last without changing it being increasing.
[ 3, 5, 6, 8, 10, 24, ]
[1, 5, 6, 8, 50, 65]
[ 4, 9, 29, ]
And now we can start enumerating the 2^6 = 64 different ways of splitting the original stream back into 2 increasing streams.
There is a famous card game in Germany called "Doppelkopf".
Usually, you play "Doppelkopf" with 4 players, but you can also play it with a table of 5 players, where one player is just watching.
(Where everyone "has the cards" once in a round, meaning everyone has the right to play the first card once every round.)
Every year, my family organizes a "Doppelkopf" tournament with 3 rounds (r).
Depending on the availabilty of my relatives, every year the number of participants varies.
Expecting a minimun of participant of 16 people, the number (n) in this experiment can rise up unlimited (as does the number of rounds r).
Naturally, my relatives do not want to be paired with someone twice, since they want to exchange gossip most efficiently!
There we have:
n - Participants
r- Rounds
t_total = n // 4 # Total Tables (round down of n)
t_5 = n % 4 # Tables of 5s
t_4 = t_total - t_5 # Tables of 4s
pos_pair = n * (n - 1) / 2 # possible pairs (n over 2)
nec_pair = (t_5 * 10 + t_4 * 6) * r # necessary pairs
I was instructed with the aim to minimize the encounters (if possible to set encounters == 1 for everyone)!
Since, I do not want to solve the problem for P{n={16, ..., 32}, r=3} (which I did for some cases), but to solve it with any given P{n∈N, r∈N} , there is a discrepancy between my abilities and the requirements for a solution!
Therefore, I would like to ask the community to help me with this problem, to solve it for any given P{n∈N, r∈N}!
And also to prove, if this problem is not solvable for any P{n∈N, r∈N}, which is given "if pos_pair < nec_pair".
Here are two solutions for P{n=20, r=3}:
which very much solves my "Doppelkopf" tournament problem:
('Best result was ', [[[1, 2, 3, 4], [5, 6, 7, 8], [9, 10, 11, 12], [13, 14, 15, 16], [17, 18, 19, 20]], [[16, 12, 8, 18], [13, 1, 5, 9], [15, 4, 17, 6], [2, 19, 7, 10], [3, 11, 20, 14]], [[14, 9, 17, 7], [13, 20, 8, 2], [5, 4, 12, 19], [6, 16, 11, 1], [15, 18, 10, 3]]])
('Best result was ', [[[1, 2, 3, 4], [5, 6, 7, 8], [9, 10, 11, 12], [13, 14, 15, 16], [17, 18, 19, 20]], [[19, 11, 13, 3], [2, 15, 9, 8], [1, 16, 18, 6], [14, 7, 17, 10], [4, 12, 20, 5]], [[17, 8, 3, 12], [20, 9, 16, 7], [15, 11, 6, 4], [2, 13, 10, 18], [1, 19, 14, 5]]])
But in order to solve this problem with an arbitrary n and r I have come to no conclusion.
In my opinion, there are three ways to go about this problem in a computational solution or approximation.
First, you can iterate about rounds, and assign every player to a
table without having collision, remembering pairs and appeareances
in total (not to exeed total rounds)
Secondly, you can iterate about tables, which seems to be helpful with participants, that are a multiple of 2 (see for P{n=16, r=5}
https://matheplanet.com/default3.html?call=viewtopic.php?topic=85206&ref=https%3A%2F%2Fwww.google.com%2F)
also remeber pairs and appearances, but mainly follow a certain
patters as described in the link, which I somehow can not scale to
other numbers!!
There is somehow a mathemathical way to descibe this procedure and conclude a solution
Even though, this is more of a mathematical question (and I don't know where to ask those questions), I am interested in the algorithmic solution!
Let's say I have the following list of lists:
x = [[1, 2, 3, 4, 5, 6, 7], # sequence 1
[6, 5, 10, 11], # sequence 2
[9, 8, 2, 3, 4, 5], # sequence 3
[12, 12, 6, 5], # sequence 4
[5, 8, 3, 4, 2], # sequence 5
[1, 5], # sequence 6
[2, 8, 8, 3, 5, 9, 1, 4, 12, 5, 6], # sequence 7
[7, 1, 7, 3, 4, 1, 2], # sequence 8
[9, 4, 12, 12, 6, 5, 1], # sequence 9
]
Essentially, for any list that contains the target number 5 (i.e., target=5) anywhere within the list, what are the top N=2 most frequently observed subsequences with length M=4?
So, the conditions are:
if target doesn't exist in the list then we ignore that list completely
if the list length is less than M then we ignore the list completely
if the list is exactly length M but target is not in the Mth position then we ignore it (but we count it if target is in the Mth position)
if the list length, L, is longer than M and target is in the i=M position(ori=M+1position, ori=M+2position, ...,i=Lposition) then we count the subsequence of lengthMwheretarget` is in the final position in the subsequence
So, using our list-of-lists example, we'd count the following subsequences:
subseqs = [[2, 3, 4, 5], # taken from sequence 1
[2, 3, 4, 5], # taken from sequence 3
[12, 12, 6, 5], # taken from sequence 4
[8, 8, 3, 5], # taken from sequence 7
[1, 4, 12, 5], # taken from sequence 7
[12, 12, 6, 5], # taken from sequence 9
]
Of course, what we want are the top N=2 subsequences by frequency. So, [2, 3, 4, 5] and [12, 12, 6, 5] are the top two most frequent sequences by count. If N=3 then all of the subsequences (subseqs) would be returned since there is a tie for third.
Important
This is super simplified but, in reality, my actual list-of-sequences
consists of a few billion lists of positive integers (between 1 and 10,000)
each list can be as short as 1 element or as long as 500 elements
N and M can be as small as 1 or as big as 100
My questions are:
Is there an efficient data structure that would allow for fast queries assuming that N and M will always be less than 100?
Are there known algorithms for performing this kind of analysis for various combinations of N and M? I've looked at suffix trees but I'd have to roll my own custom version to even get close to what I need.
For the same dataset, I need to repeatedly query the dataset for various values or different combinations of target, N, and M (where target <= 10,000, N <= 100 and `M <= 100). How can I do this efficiently?
Extending on my comment. Here is a sketch how you could approach this using an out-of-the-box suffix array:
1) reverse and concatenate your lists with a stop symbol (I used 0 here).
[7, 6, 5, 4, 3, 2, 1, 0, 11, 10, 5, 6, 0, 5, 4, 3, 2, 8, 9, 0, 5, 6, 12, 12, 0, 2, 4, 3, 8, 5, 0, 5, 1, 0, 6, 5, 12, 4, 1, 9, 5, 3, 8, 8, 2, 0, 2, 1, 4, 3, 7, 1, 7, 0, 1, 5, 6, 12, 12, 4, 9]
2) Build a suffix array
[53, 45, 24, 30, 12, 19, 33, 7, 32, 6, 47, 54, 51, 38, 44, 5, 46, 25, 16, 4, 15, 49, 27, 41, 37, 3, 14, 48, 26, 59, 29, 31, 40, 2, 13, 10, 20, 55, 35, 11, 1, 34, 21, 56, 52, 50, 0, 43, 28, 42, 17, 18, 39, 60, 9, 8, 23, 36, 58, 22, 57]
3) Build the LCP array. The LCP array will tell you how many numbers a suffix has in common with its neighbour in the suffix array. However, you need to stop counting when you encounter a stop symbol
[0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 0, 1, 2, 1, 1, 0, 2, 1, 1, 2, 0, 1, 3, 2, 2, 1, 0, 1, 1, 1, 4, 1, 2, 4, 1, 0, 1, 2, 1, 3, 0, 1, 1, 0, 1, 1, 1, 0, 1, 1, 0, 0, 0, 1, 2, 1, 2, 0]
4) When a query comes in (target = 5, M= 4) you search for the first occurence of your target in the suffix array and scan the corresponding LCP-array until the starting number of suffixes changes. Below is the part of the LCP array that corresponds to all suffixes starting with 5.
[..., 1, 1, 1, 4, 1, 2, 4, 1, 0, ...]
This tells you that there are two sequences of length 4 that occur two times. Brushing over some details using the indexes you can find the sequences and revert them back to get your final results.
Complexity
Building up the suffix array is O(n) where n is the total number of elements in all lists and O(n) space
Building the LCP array is also O(n) in both time and space
Searching a target number in the suffix is O(log n) in average
The cost of scanning through the relevant subsequences is linear in the number of times the target occurs. Which should be 1/10000 on average according to your given parameters.
The first two steps happen offline. Querying is technically O(n) (due to step 4) but with a small constant (0.0001).
I'm working on a "provably fair" site where let's say X participants enter into a drawing and we need to pick first 1 overall winner, but then ideally we also want to pick N sub-winners out of the X total.
(for the curious, the SHA-256 Hash will be the merkle tree root of a Bitcoin block at a pre-specified time)
So, given a SHA-256 hash, how do we generate N random numbers?
I think I know how to generate 1 random number (within ruby's Fixnum range). According to this article: http://patshaughnessy.net/2014/1/9/how-big-is-a-bignum
The maximum Fixnum integer is: 4611686018427387903
Let's pluck the first Y characters of the SHA-256 hash. We can generate one instead of relying on a Bitcoin merkle root with:
d = Digest::SHA256.hexdigest('hello')
> "2cf24dba5fb0a30e26e83b2ac5b9e29e1b161e5c1fa7425e73043362938b9824"
Let's take the first 6 characters, or: 2cf24d
Convert this to base 10:
'2cf24d'.to_i(16)
> 2945613
We now have a unique Fixnum based on our merkle root.
With X participants, let's say 17, we decide the winner with:
2945613 % 17
> 6
So assuming all entries know their order of entry, the sixth entrant can prove that they should be the winner.
Now -- what would be the best way to similarly pick N sub-winners? Let's say each of these entrants should get a smaller but still somewhat valuable prize.
Why not just use the hash for the seed?
[*1..17].shuffle(random: Random.new(0x2cf24d))
# => [15, 5, 9, 7, 14, 3, 16, 12, 2, 1, 17, 4, 6, 13, 11, 10, 8]
[*1..17].shuffle(random: Random.new(0x2cf24d))
# => [15, 5, 9, 7, 14, 3, 16, 12, 2, 1, 17, 4, 6, 13, 11, 10, 8]
EDIT: This is dependent on Ruby version though - I believe shuffle is different between JRuby and MRI, even though Random produces the same sequence. You could circumvent this by implementing shuffle yourself. See this question for more details. This workaround works consistently for me in both JRuby and MRI:
r = Random.new(0x2cf24d)
[*1..17].sort_by { r.rand }
# => [14, 11, 4, 10, 1, 3, 9, 13, 16, 17, 12, 5, 8, 2, 6, 7, 15]
r = Random.new(0x2cf24d)
[*1..17].sort_by { r.rand }
# => [14, 11, 4, 10, 1, 3, 9, 13, 16, 17, 12, 5, 8, 2, 6, 7, 15]
This question already has answers here:
Closed 10 years ago.
Possible Duplicate:
Find the min number in all contiguous subarrays of size l of a array of size n
I have a (large) array of numeric data (size N) and would like to compute an array of running maximums with a fixed window size w.
More directly, I can define a new array out[k-w+1] = max{data[k-w+1,...,k]} for k >= w-1 (this assumes 0-based arrays, as in C++).
Is there a better way to do this than N log(w)?
[I'm hoping there should be a linear one in N without dependence on w, like for moving average, but cannot find it. For N log(w) I think there is a way to manage with a sorted data structure which will do insert(), delete() and extract_max() altogether in log(w) or less on a structure of size w -- like a sorted binary tree, for example].
Thank you very much.
There is indeed an algorithm that can do this in O(N) time with no dependence on the window size w. The idea is to use a clever data structure that supports the following operations:
Enqueue, which adds a new element to the structure,
Dequeue, which removes the oldest element from the structure, and
Find-max, which returns (but does not remove) the minimum element from the structure.
This is essentially a queue data structure that supports access (but not removal) of the maximum element. Amazingly, as seen in this earlier question, it is possible to implement this data structure such that each of these operations runs in amortized O(1) time. As a result, if you use this structure to enqueue w elements, then continuously dequeue and enqueue another element into the structure while calling find-max as needed, it will take only O(n + Q) time, where Q is the number of queries you make. If you only care about the minimum of each window once, this ends up being O(n), with no dependence on the window size.
Hope this helps!
I'll demonstrate how to do it with the list:
L = [21, 17, 16, 7, 3, 9, 11, 18, 19, 5, 10, 23, 20, 15, 4, 14, 1, 2, 22, 13, 8, 12, 6]
with length N=23 and W = 4.
Make two new copies of your list:
L1 = [21, 17, 16, 7, 3, 9, 11, 18, 19, 5, 10, 23, 20, 15, 4, 14, 1, 2, 22, 13, 8, 12, 6]
L2 = [21, 17, 16, 7, 3, 9, 11, 18, 19, 5, 10, 23, 20, 15, 4, 14, 1, 2, 22, 13, 8, 12, 6]
Loop from i=0 to N-1. If i is not divisible by W, then replace L1[i] with max(L1[i],L1[i-1]).
L1 = [21, 21, 21, 21, | 3, 9, 11, 18, | 19, 19, 19, 23 | 20, 20, 20, 20 | 1, 2, 22, 22 | 8, 12, 12]
Loop from i=N-2 to0. If i+1 is not divisible by W, then replace L2[i] with max(L2[i], L2[i+1]).
L2 = [21, 17, 16, 7 | 18, 18, 18, 18 | 23, 23, 23, 23 | 20, 15, 14, 14 | 22, 22, 22, 13 | 12, 12, 6]
Make a list L3 of length N + 1 - W, so that L3[i] = max(L2[i], L1[i + W - 1])
L3 = [21, 17, 16, 11 | 18, 19, 19, 19 | 23, 23, 23, 23 | 20, 15, 14, 22 | 22, 22, 22, 13]
Then this list L3 is the moving maxima you seek, L2[i] is the maximum of the range between i and the next vertical line, while l1[i + W - 1] is the maximum of the range between the vertical line and i + W - 1.