Let a1,...,an be a sequence of real numbers. Let m be the minimum of the sequence, and let M be the maximum of the sequence.
I proved that there exists 2 elements in the sequence, x,y, such that |x-y|<=(M-m)/n.
Now, is there a way to find an algorithm that finds such 2 elements in time complexity of O(n)?
I thought about sorting the sequence, but since I dont know anything about M I cannot use radix/bucket or any other linear time algorithm that I'm familier with.
I'd appreciate any idea.
Thanks in advance.
First find out n, M, m. If not already given they can be determined in O(n).
Then create a memory storage of n+1 elements; we will use the storage for n+1 buckets with width w=(M-m)/n.
The buckets cover the range of values equally: Bucket 1 goes from [m; m+w[, Bucket 2 from [m+w; m+2*w[, Bucket n from [m+(n-1)*w; m+n*w[ = [M-w; M[, and the (n+1)th bucket from [M; M+w[.
Now we go once through all the values and sort them into the buckets according to the assigned intervals. There should be at a maximum 1 element per bucket. If the bucket is already filled, it means that the elements are closer together than the boundaries of the half-open interval, e.g. we found elements x, y with |x-y| < w = (M-m)/n.
If no such two elements are found, afterwards n buckets of n+1 total buckets are filled with one element. And all those elements are sorted.
We once more go through all the buckets and compare the distance of the content of neighbouring buckets only, whether there are two elements, which fulfil the condition.
Due to the width of the buckets, the condition cannot be true for buckets, which are not adjoining: For those the distance is always |x-y| > w.
(The fulfilment of the last inequality in 4. is also the reason, why the interval is half-open and cannot be closed, and why we need n+1 buckets instead of n. An alternative would be, to use n buckets and make the now last bucket a special case with [M; M+w]. But O(n+1)=O(n) and using n+1 steps is preferable to special casing the last bucket.)
The running time is O(n) for step 1, 0 for step 2 - we actually do not do anything there, O(n) for step 3 and O(n) for step 4, as there is only 1 element per bucket. Altogether O(n).
This task shows, that either sorting of elements, which are not close together or coarse sorting without considering fine distances can be done in O(n) instead of O(n*log(n)). It has useful applications. Numbers on computers are discrete, they have a finite precision. I have sucessfuly used this sorting method for signal-processing / fast sorting in real-time production code.
About #Damien 's remark: The real threshold of (M-m)/(n-1) is provably true for every such sequence. I assumed in the answer so far the sequence we are looking at is a special kind, where the stronger condition is true, or at least, for all sequences, if the stronger condition was true, we would find such elements in O(n).
If this was a small mistake of the OP instead (who said to have proven the stronger condition) and we should find two elements x, y with |x-y| <= (M-m)/(n-1) instead, we can simplify:
-- 3. We would do steps 1 to 3 like above, but with n buckets and the bucket width set to w = (M-m)/(n-1). The bucket n now goes from [M; M+w[.
For step 4 we would do the following alternative:
4./alternative: n buckets are filled with one element each. The element at bucket n has to be M and is at the left boundary of the bucket interval. The distance of this element y = M to the element x in the n-1th bucket for every such possible element x in the n-1thbucket is: |M-x| <= w = (M-m)/(n-1), so we found x and y, which fulfil the condition, q.e.d.
First note that the real threshold should be (M-m)/(n-1).
The first step is to calculate the min m and max M elements, in O(N).
You calculate the mid = (m + M)/2value.
You concentrate the value less than mid at the beginning, and more than mid at the end of he array.
You select the part with the largest number of elements and you iterate until very few numbers are kept.
If both parts have the same number of elements, you can select any of them. If the remaining part has much more elements than n/2, then in order to maintain a O(n) complexity, you can keep onlyn/2 + 1 of them, as the goal is not to find the smallest difference, but one difference small enough only.
As indicated in a comment by #btilly, this solution could fail in some cases, for example with an input [0, 2.1, 2.9, 5]. For that, it is needed to calculate the max value of the left hand, and the min value of the right hand, and to test if the answer is not right_min - left_max. This doesn't change the O(n) complexity, even if the solution becomes less elegant.
Complexity of the search procedure: O(n) + O(n/2) + O(n/4) + ... + O(2) = O(2n) = O(n).
Damien is correct in his comment that the correct results is that there must be x, y such that |x-y| <= (M-m)/(n-1). If you have the sequence [0, 1, 2, 3, 4] you have 5 elements, but no two elements are closer than (M-m)/n = (4-0)/5 = 4/5.
With the right threshold, the solution is easy - find M and m by scanning through the input once, and then bucket the input into (n-1) buckets of size (M-m)/(n-1), putting values that are on the boundaries of a pair of buckets into both buckets. At least one bucket must have two values in it by the pigeon-hole principle.
Related
Given two sorted arrays of numbers, we want to find the pair with the kth largest possible sum. (A pair is one element from the first array and one element from the second array). For example, with arrays
[2, 3, 5, 8, 13]
[4, 8, 12, 16]
The pairs with largest sums are
13 + 16 = 29
13 + 12 = 25
8 + 16 = 24
13 + 8 = 21
8 + 12 = 20
So the pair with the 4th largest sum is (13, 8). How to find the pair with the kth largest possible sum?
Also, what is the fastest algorithm? The arrays are already sorted and sizes M and N.
I am already aware of the O(Klogk) solution , using Max-Heap given here .
It also is one of the favorite Google interview question , and they demand a O(k) solution .
I've also read somewhere that there exists a O(k) solution, which i am unable to figure out .
Can someone explain the correct solution with a pseudocode .
P.S.
Please DON'T post this link as answer/comment.It DOESN'T contain the answer.
I start with a simple but not quite linear-time algorithm. We choose some value between array1[0]+array2[0] and array1[N-1]+array2[N-1]. Then we determine how many pair sums are greater than this value and how many of them are less. This may be done by iterating the arrays with two pointers: pointer to the first array incremented when sum is too large and pointer to the second array decremented when sum is too small. Repeating this procedure for different values and using binary search (or one-sided binary search) we could find Kth largest sum in O(N log R) time, where N is size of the largest array and R is number of possible values between array1[N-1]+array2[N-1] and array1[0]+array2[0]. This algorithm has linear time complexity only when the array elements are integers bounded by small constant.
Previous algorithm may be improved if we stop binary search as soon as number of pair sums in binary search range decreases from O(N2) to O(N). Then we fill auxiliary array with these pair sums (this may be done with slightly modified two-pointers algorithm). And then we use quickselect algorithm to find Kth largest sum in this auxiliary array. All this does not improve worst-case complexity because we still need O(log R) binary search steps. What if we keep the quickselect part of this algorithm but (to get proper value range) we use something better than binary search?
We could estimate value range with the following trick: get every second element from each array and try to find the pair sum with rank k/4 for these half-arrays (using the same algorithm recursively). Obviously this should give some approximation for needed value range. And in fact slightly improved variant of this trick gives range containing only O(N) elements. This is proven in following paper: "Selection in X + Y and matrices with sorted rows and columns" by A. Mirzaian and E. Arjomandi. This paper contains detailed explanation of the algorithm, proof, complexity analysis, and pseudo-code for all parts of the algorithm except Quickselect. If linear worst-case complexity is required, Quickselect may be augmented with Median of medians algorithm.
This algorithm has complexity O(N). If one of the arrays is shorter than other array (M < N) we could assume that this shorter array is extended to size N with some very small elements so that all calculations in the algorithm use size of the largest array. We don't actually need to extract pairs with these "added" elements and feed them to quickselect, which makes algorithm a little bit faster but does not improve asymptotic complexity.
If k < N we could ignore all the array elements with index greater than k. In this case complexity is equal to O(k). If N < k < N(N-1) we just have better complexity than requested in OP. If k > N(N-1), we'd better solve the opposite problem: k'th smallest sum.
I uploaded simple C++11 implementation to ideone. Code is not optimized and not thoroughly tested. I tried to make it as close as possible to pseudo-code in linked paper. This implementation uses std::nth_element, which allows linear complexity only on average (not worst-case).
A completely different approach to find K'th sum in linear time is based on priority queue (PQ). One variation is to insert largest pair to PQ, then repeatedly remove top of PQ and instead insert up to two pairs (one with decremented index in one array, other with decremented index in other array). And take some measures to prevent inserting duplicate pairs. Other variation is to insert all possible pairs containing largest element of first array, then repeatedly remove top of PQ and instead insert pair with decremented index in first array and same index in second array. In this case there is no need to bother about duplicates.
OP mentions O(K log K) solution where PQ is implemented as max-heap. But in some cases (when array elements are evenly distributed integers with limited range and linear complexity is needed only on average, not worst-case) we could use O(1) time priority queue, for example, as described in this paper: "A Complexity O(1) Priority Queue for Event Driven Molecular Dynamics Simulations" by Gerald Paul. This allows O(K) expected time complexity.
Advantage of this approach is a possibility to provide first K elements in sorted order. Disadvantages are limited choice of array element type, more complex and slower algorithm, worse asymptotic complexity: O(K) > O(N).
EDIT: This does not work. I leave the answer, since apparently I am not the only one who could have this kind of idea; see the discussion below.
A counter-example is x = (2, 3, 6), y = (1, 4, 5) and k=3, where the algorithm gives 7 (3+4) instead of 8 (3+5).
Let x and y be the two arrays, sorted in decreasing order; we want to construct the K-th largest sum.
The variables are: i the index in the first array (element x[i]), j the index in the second array (element y[j]), and k the "order" of the sum (k in 1..K), in the sense that S(k)=x[i]+y[j] will be the k-th greater sum satisfying your conditions (this is the loop invariant).
Start from (i, j) equal to (0, 0): clearly, S(1) = x[0]+y[0].
for k from 1 to K-1, do:
if x[i+1]+ y[j] > x[i] + y[j+1], then i := i+1 (and j does not change) ; else j:=j+1
To see that it works, consider you have S(k) = x[i] + y[j]. Then, S(k+1) is the greatest sum which is lower (or equal) to S(k), and such as at least one element (i or j) changes. It is not difficult to see that exactly one of i or j should change.
If i changes, the greater sum you can construct which is lower than S(k) is by setting i=i+1, because x is decreasing and all the x[i'] + y[j] with i' < i are greater than S(k). The same holds for j, showing that S(k+1) is either x[i+1] + y[j] or x[i] + y[j+1].
Therefore, at the end of the loop you found the K-th greater sum.
tl;dr: If you look ahead and look behind at each iteration, you can start with the end (which is highest) and work back in O(K) time.
Although the insight underlying this approach is, I believe, sound, the code below is not quite correct at present (see comments).
Let's see: first of all, the arrays are sorted. So, if the arrays are a and b with lengths M and N, and as you have arranged them, the largest items are in slots M and N respectively, the largest pair will always be a[M]+b[N].
Now, what's the second largest pair? It's going to have perhaps one of {a[M],b[N]} (it can't have both, because that's just the largest pair again), and at least one of {a[M-1],b[N-1]}. BUT, we also know that if we choose a[M-1]+b[N-1], we can make one of the operands larger by choosing the higher number from the same list, so it will have exactly one number from the last column, and one from the penultimate column.
Consider the following two arrays: a = [1, 2, 53]; b = [66, 67, 68]. Our highest pair is 53+68. If we lose the smaller of those two, our pair is 68+2; if we lose the larger, it's 53+67. So, we have to look ahead to decide what our next pair will be. The simplest lookahead strategy is simply to calculate the sum of both possible pairs. That will always cost two additions, and two comparisons for each transition (three because we need to deal with the case where the sums are equal);let's call that cost Q).
At first, I was tempted to repeat that K-1 times. BUT there's a hitch: the next largest pair might actually be the other pair we can validly make from {{a[M],b[N]}, {a[M-1],b[N-1]}. So, we also need to look behind.
So, let's code (python, should be 2/3 compatible):
def kth(a,b,k):
M = len(a)
N = len(b)
if k > M*N:
raise ValueError("There are only %s possible pairs; you asked for the %sth largest, which is impossible" % M*N,k)
(ia,ib) = M-1,N-1 #0 based arrays
# we need this for lookback
nottakenindices = (0,0) # could be any value
nottakensum = float('-inf')
for i in range(k-1):
optionone = a[ia]+b[ib-1]
optiontwo = a[ia-1]+b[ib]
biggest = max((optionone,optiontwo))
#first deal with look behind
if nottakensum > biggest:
if optionone == biggest:
newnottakenindices = (ia,ib-1)
else: newnottakenindices = (ia-1,ib)
ia,ib = nottakenindices
nottakensum = biggest
nottakenindices = newnottakenindices
#deal with case where indices hit 0
elif ia <= 0 and ib <= 0:
ia = ib = 0
elif ia <= 0:
ib-=1
ia = 0
nottakensum = float('-inf')
elif ib <= 0:
ia-=1
ib = 0
nottakensum = float('-inf')
#lookahead cases
elif optionone > optiontwo:
#then choose the first option as our next pair
nottakensum,nottakenindices = optiontwo,(ia-1,ib)
ib-=1
elif optionone < optiontwo: # choose the second
nottakensum,nottakenindices = optionone,(ia,ib-1)
ia-=1
#next two cases apply if options are equal
elif a[ia] > b[ib]:# drop the smallest
nottakensum,nottakenindices = optiontwo,(ia-1,ib)
ib-=1
else: # might be equal or not - we can choose arbitrarily if equal
nottakensum,nottakenindices = optionone,(ia,ib-1)
ia-=1
#+2 - one for zero-based, one for skipping the 1st largest
data = (i+2,a[ia],b[ib],a[ia]+b[ib],ia,ib)
narrative = "%sth largest pair is %s+%s=%s, with indices (%s,%s)" % data
print (narrative) #this will work in both versions of python
if ia <= 0 and ib <= 0:
raise ValueError("Both arrays exhausted before Kth (%sth) pair reached"%data[0])
return data, narrative
For those without python, here's an ideone: http://ideone.com/tfm2MA
At worst, we have 5 comparisons in each iteration, and K-1 iterations, which means that this is an O(K) algorithm.
Now, it might be possible to exploit information about differences between values to optimise this a little bit, but this accomplishes the goal.
Here's a reference implementation (not O(K), but will always work, unless there's a corner case with cases where pairs have equal sums):
import itertools
def refkth(a,b,k):
(rightia,righta),(rightib,rightb) = sorted(itertools.product(enumerate(a),enumerate(b)), key=lamba((ia,ea),(ib,eb):ea+eb)[k-1]
data = k,righta,rightb,righta+rightb,rightia,rightib
narrative = "%sth largest pair is %s+%s=%s, with indices (%s,%s)" % data
print (narrative) #this will work in both versions of python
return data, narrative
This calculates the cartesian product of the two arrays (i.e. all possible pairs), sorts them by sum, and takes the kth element. The enumerate function decorates each item with its index.
The max-heap algorithm in the other question is simple, fast and correct. Don't knock it. It's really well explained too. https://stackoverflow.com/a/5212618/284795
Might be there isn't any O(k) algorithm. That's okay, O(k log k) is almost as fast.
If the last two solutions were at (a1, b1), (a2, b2), then it seems to me there are only four candidate solutions (a1-1, b1) (a1, b1-1) (a2-1, b2) (a2, b2-1). This intuition could be wrong. Surely there are at most four candidates for each coordinate, and the next highest is among the 16 pairs (a in {a1,a2,a1-1,a2-1}, b in {b1,b2,b1-1,b2-1}). That's O(k).
(No it's not, still not sure whether that's possible.)
[2, 3, 5, 8, 13]
[4, 8, 12, 16]
Merge the 2 arrays and note down the indexes in the sorted array. Here is the index array looks like (starting from 1 not 0)
[1, 2, 4, 6, 8]
[3, 5, 7, 9]
Now start from end and make tuples. sum the elements in the tuple and pick the kth largest sum.
public static List<List<Integer>> optimization(int[] nums1, int[] nums2, int k) {
// 2 * O(n log(n))
Arrays.sort(nums1);
Arrays.sort(nums2);
List<List<Integer>> results = new ArrayList<>(k);
int endIndex = 0;
// Find the number whose square is the first one bigger than k
for (int i = 1; i <= k; i++) {
if (i * i >= k) {
endIndex = i;
break;
}
}
// The following Iteration provides at most endIndex^2 elements, and both arrays are in ascending order,
// so k smallest pairs must can be found in this iteration. To flatten the nested loop, refer
// 'https://stackoverflow.com/questions/7457879/algorithm-to-optimize-nested-loops'
for (int i = 0; i < endIndex * endIndex; i++) {
int m = i / endIndex;
int n = i % endIndex;
List<Integer> item = new ArrayList<>(2);
item.add(nums1[m]);
item.add(nums2[n]);
results.add(item);
}
results.sort(Comparator.comparing(pair->pair.get(0) + pair.get(1)));
return results.stream().limit(k).collect(Collectors.toList());
}
Key to eliminate O(n^2):
Avoid cartesian product(or 'cross join' like operation) of both arrays, which means flattening the nested loop.
Downsize iteration over the 2 arrays.
So:
Sort both arrays (Arrays.sort offers O(n log(n)) performance according to Java doc)
Limit the iteration range to the size which is just big enough to support k smallest pairs searching.
First let me phrase the proper question:
Q: There is a file containing more than a million points (x,y) each of which represents a star. There is a planet earth at (a,b). Now, the task is to build an algorithm that would return the 100 closest stars to earth. What would be the time and space complexities of your algorithm.
This question has been asked many times in various interviews. I tried looking up the answers but could not find a satisfactory one.
One way to do it which I thought might be using a max heap of size 100. Calculate distances for each star and check if the distance is lesser than the root of the max heap. If yes, replace it with the root and call heapify.
Any other better/faster answers?
P.S: This is not a homework question.
You can actually do this in time O(n) and space O(k), where k is the number of closest points that you want, by using a very clever trick.
The selection problem is as follows: Given an array of elements and some index i, rearrange the elements of the array such that the ith element is in the right place, all elements smaller than the ith element are to the left, and all elements greater than the ith element are to the right. For example, given the array
40 10 00 30 20
If I tried to select based on index 2 (zero-indexed), one result might be
10 00 20 40 30
Since the element at index 2 (20) is in the right place, the elements to the left are smaller than 20, and the elements to the right are greater than 20.
It turns out that since this is a less strict requirement than actually sorting the array, it's possible to do this in time O(n), where n is the number of elements of the array. Doing so requires some complex algorithms like the median-of-medians algorithm, but is indeed O(n) time.
So how do you use this here? One option is to load all n elements from the file into an array, then use the selection algorithm to select the top k in O(n) time and O(n) space (here, k = 100).
But you can actually do better than this! For any constant k that you'd like, maintain a buffer of 2k elements. Load 2k elements from the file into the array, then use the selection algorithm to rearrange it so that the smallest k elements are in the left half of the array and the largest are in the right, then discard the largest k elements (they can't be any of the k closest points). Now, load k more elements from the file into the buffer and do this selection again, and repeat this until you've processed every line of the file. Each time you do a selection you discard the largest k elements in the buffer and retain the k closest points you've seen so far. Consequently, at the very end, you can select the top k elements one last time and find the top k.
What's the complexity of the new approach? Well, you're using O(k) memory for the buffer and the selection algorithm. You end up calling select on a buffer of size O(k) a total of O(n / k) times, since you call select after reading k new elements. Since select on a buffer of size O(k) takes time O(k), the total runtime here is O(n + k). If k = O(n) (a reasonable assumption), this takes time O(n), space O(k).
Hope this helps!
To elaborate on the MaxHeap solution you would build a max-heap with the first k elements from the file ( k = 100 in this case ).
The key for the max-heap would be its distance from Earth (a,b). Distance between 2 points on a 2d plane can be calculated using:
dist = (x1,y1) to (x2,y2) = square_root((x2 - x1)^2 + (y2 - y1)^2);
This would take O(k) time to construct. For every subsequent element from k to n. ie (n - k) elements you need to fetch its distance from earth and compare it with the top of max-heap. If the new element to be inserted is closer to earth than the top of the max-heap, replace the top of the max-heap and call heapify on the new root of the heap.
This would take O((n-k)logk) time to complete.
Finally we would be left with just the k elements in the max-heap. You can call heapify k times to return all these k elements. This is another O(klogk).
Overall time complexity would be O(k + (n-k)logk + klogk).
It's a famous question and there has been lot's of solution for that:
http://en.wikipedia.org/wiki/K-nearest_neighbor_algorithm
if you did not find it useful, there are some other resources such as Rurk's computational geometry book.
Your algorithm is correct. Just remember that time complexity of your program is O(n . log 100 ) = O(n), unless number of closest points to find can vary.
import sys,os,csv
iFile=open('./file_copd.out','rU')
earth = [0,0]
##getDistance return distance given two stars
def getDistance(star1,star2):
return sqrt((star1[0]-star2[0])**2 +(star1[1]-star2[1])**2 )
##diction dict_galaxy looks like this {key,distance} key is the seq assign to each star, value is a list [distance,its cordinance]
##{1,[distance1,[x,y]];2,[distance2,[x,y]]}
dict_galaxy={}
#list_galaxy=[]
count = 0
sour=iFile.readlines()
for line in sour:
star=line.split(',') ##Star is a list [x,y]
dict_galaxy[count]=[getDistance(earth,star),star]
count++
###Now sort this dictionary based on their distance, and return you a list of keys.
list_sorted_key = sorted(dict_galaxy,key=lambda x:dict_galaxy[x][0])
print 'is this what you want %s'%(list_sorted_key[:100].to_s)
iFile.close()
I was given the following question in an algorithms book:
Suppose a merge sort is implemented to split a file at a random position, rather then exactly in the middle. How many comparisons would be used by such method to sort n elements on average?
Thanks.
To guide you to the answer, consider these more specific questions:
Assume the split is always at 10%, or 25%, or 75%, or 90%. In each case: what's the impact on recursion depths? How many comparisons need to be per recursion level?
I'm partially agree with #Armen, they should be comparable.
But: consider the case when they are split in the middle. To merge two lists of lengths n we would need 2*n - 1 comparations (sometimes less, but we'll consider it fixed for simplicity), each of them producing the next value. There would be log2(n) levels of merges, that gives us approximately n*log2(n) comparations.
Now considering the random-split case: The maximum number of comparations needed to merge a list of length n1 with one of length n2 will be n1 + n2 - 1. Howerer, the average number will be close to it, because even for the most unhappy split 1 and n-1 we'll need an average of n/2 comparations. So we can consider that the cost of merging per level will be the same as in even case.
The difference is that in random case the number of levels will be larger, and we can consider that n for next level would be max(n1, n2) instead of n/2. This max(n1, n2) will tend to be 3*n/4, that gives us the approximate formula
n*log43(n) // where log43 is log in base 4/3
that gives us
n * log2(n) / log2(4/3) ~= 2.4 * n * log2(n)
This result is still larger than the correct one because we ignored that the small list will have fewer levels, but it should be close enough. I suppose that the correct answer will be the number of comparations on average will double
You can get an upper bound of 2n * H_{n - 1} <= 2n ln n using the fact that merging two lists of total length n costs at most n comparisons. The analysis is similar to that of randomized quicksort (see http://www.cs.cmu.edu/afs/cs/academic/class/15451-s07/www/lecture_notes/lect0123.pdf).
First, suppose we split a list of length n into 2 lists L and R. We will charge the first element of R for a comparison against all of the elements of L, and the last element of L for a comparison against all elements of R. Even though these may not be the exact comparisons that are executed, the total number of comparisons we are charging for is n as required.
This handles one level of recursion, but what about the rest? We proceed by concentrating only on the "right-to-left" comparisons that occur between the first element of R and every element of L at all levels of recursion (by symmetry, this will be half the actual expected total). The probability that the jth element is compared to the ith element is 1/(j - i) where j > i. To see this, note that element j is compared with element i exactly when it is the first element chosen as a "splitting element" from among the set {i + 1,..., j}. That is, elements i and j are split into two lists as soon as the list they are in is split at some element from {i + 1,..., j}, and element j is charged for a comparison with i exactly when element j is the element that is chosen from this set.
Thus, the expected total number of comparisons involving j is at most H_n (i.e., 1 + 1/2 + 1/3..., where the number of terms is at most n - 1). Summing across all possible j gives n * H_{n - 1}. This only counted "right-to-left" comparisons, so the final upper bound is 2n * H_{n - 1}.
We've got some nonnegative numbers. We want to find the pair with maximum gcd. actually this maximum is more important than the pair!
For example if we have:
2 4 5 15
gcd(2,4)=2
gcd(2,5)=1
gcd(2,15)=1
gcd(4,5)=1
gcd(4,15)=1
gcd(5,15)=5
The answer is 5.
You can use the Euclidean Algorithm to find the GCD of two numbers.
while (b != 0)
{
int m = a % b;
a = b;
b = m;
}
return a;
If you want an alternative to the obvious algorithm, then assuming your numbers are in a bounded range, and you have plenty of memory, you can beat O(N^2) time, N being the number of values:
Create an array of a small integer type, indexes 1 to the max input. O(1)
For each value, increment the count of every element of the index which is a factor of the number (make sure you don't wraparound). O(N).
Starting at the end of the array, scan back until you find a value >= 2. O(1)
That tells you the max gcd, but doesn't tell you which pair produced it. For your example input, the computed array looks like this:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
4 2 1 1 2 0 0 0 0 0 0 0 0 0 1
I don't know whether this is actually any faster for the inputs you have to handle. The constant factors involved are large: the bound on your values and the time to factorise a value within that bound.
You don't have to factorise each value - you could use memoisation and/or a pregenerated list of primes. Which gives me the idea that if you are memoising the factorisation, you don't need the array:
Create an empty set of int, and a best-so-far value 1.
For each input integer:
if it's less than or equal to best-so-far, continue.
check whether it's in the set. If so, best-so-far = max(best-so-far, this-value), continue. If not:
add it to the set
repeat for all of its factors (larger than best-so-far).
Add/lookup in a set could be O(log N), although it depends what data structure you use. Each value has O(f(k)) factors, where k is the max value and I can't remember what the function f is...
The reason that you're finished with a value as soon as you encounter it in the set is that you've found a number which is a common factor of two input values. If you keep factorising, you'll only find smaller such numbers, which are not interesting.
I'm not quite sure what the best way is to repeat for the larger factors. I think in practice you might have to strike a balance: you don't want to do them quite in decreasing order because it's awkward to generate ordered factors, but you also don't want to actually find all the factors.
Even in the realms of O(N^2), you might be able to beat the use of the Euclidean algorithm:
Fully factorise each number, storing it as a sequence of exponents of primes (so for example 2 is {1}, 4 is {2}, 5 is {0, 0, 1}, 15 is {0, 1, 1}). Then you can calculate gcd(a,b) by taking the min value at each index and multiplying them back out. No idea whether this is faster than Euclid on average, but it might be. Obviously it uses a load more memory.
The optimisations I can think of is
1) start with the two biggest numbers since they are likely to have most prime factors and thus likely to have the most shared prime factors (and thus the highest GCD).
2) When calculating the GCDs of other pairs you can stop your Euclidean algorithm loop if you get below your current greatest GCD.
Off the top of my head I can't think of a way that you can work out the greatest GCD of a pair without trying to work out each pair individually (and optimise a bit as above).
Disclaimer: I've never looked at this problem before and the above is off the top of my head. There may be better ways and I may be wrong. I'm happy to discuss my thoughts in more length if anybody wants. :)
There is no O(n log n) solution to this problem in general. In fact, the worst case is O(n^2) in the number of items in the list. Consider the following set of numbers:
2^20 3^13 5^9 7^2*11^4 7^4*11^3
Only the GCD of the last two is greater than 1, but the only way to know that from looking at the GCDs is to try out every pair and notice that one of them is greater than 1.
So you're stuck with the boring brute-force try-every-pair approach, perhaps with a couple of clever optimizations to avoid doing needless work when you've already found a large GCD (while making sure that you don't miss anything).
With some constraints, e.g the numbers in the array are within a given range, say 1-1e7, it is doable in O(NlogN) / O(MAX * logMAX), where MAX is the maximum possible value in A.
Inspired from the sieve algorithm, and came across it in a Hackerrank Challenge -- there it is done for two arrays. Check their editorial.
find min(A) and max(A) - O(N)
create a binary mask, to mark which elements of A appear in the given range, for O(1) lookup; O(N) to build; O(MAX_RANGE) storage.
for every number a in the range (min(A), max(A)):
for aa = a; aa < max(A); aa += a:
if aa in A, increment a counter for aa, and compare it to current max_gcd, if counter >= 2 (i.e, you have two numbers divisible by aa);
store top two candidates for each GCD candidate.
could also ignore elements which are less than current max_gcd;
Previous answer:
Still O(N^2) -- sort the array; should eliminate some of the unnecessary comparisons;
max_gcd = 1
# assuming you want pairs of distinct elements.
sort(a) # assume in place
for ii = n - 1: -1 : 0 do
if a[ii] <= max_gcd
break
for jj = ii - 1 : -1 :0 do
if a[jj] <= max_gcd
break
current_gcd = GCD(a[ii], a[jj])
if current_gcd > max_gcd:
max_gcd = current_gcd
This should save some unnecessary computation.
There is a solution that would take O(n):
Let our numbers be a_i. First, calculate m=a_0*a_1*a_2*.... For each number a_i, calculate gcd(m/a_i, a_i). The number you are looking for is the maximum of these values.
I haven't proved that this is always true, but in your example, it works:
m=2*4*5*15=600,
max(gcd(m/2,2), gcd(m/4,4), gcd(m/5,5), gcd(m/15,15))=max(2, 2, 5, 5)=5
NOTE: This is not correct. If the number a_i has a factor p_j repeated twice, and if two other numbers also contain this factor, p_j, then you get the incorrect result p_j^2 insted of p_j. For example, for the set 3, 5, 15, 25, you get 25 as the answer instead of 5.
However, you can still use this to quickly filter out numbers. For example, in the above case, once you determine the 25, you can first do the exhaustive search for a_3=25 with gcd(a_3, a_i) to find the real maximum, 5, then filter out gcd(m/a_i, a_i), i!=3 which are less than or equal to 5 (in the example above, this filters out all others).
Added for clarification and justification:
To see why this should work, note that gcd(a_i, a_j) divides gcd(m/a_i, a_i) for all j!=i.
Let's call gcd(m/a_i, a_i) as g_i, and max(gcd(a_i, a_j),j=1..n, j!=i) as r_i. What I say above is g_i=x_i*r_i, and x_i is an integer. It is obvious that r_i <= g_i, so in n gcd operations, we get an upper bound for r_i for all i.
The above claim is not very obvious. Let's examine it a bit deeper to see why it is true: the gcd of a_i and a_j is the product of all prime factors that appear in both a_i and a_j (by definition). Now, multiply a_j with another number, b. The gcd of a_i and b*a_j is either equal to gcd(a_i, a_j), or is a multiple of it, because b*a_j contains all prime factors of a_j, and some more prime factors contributed by b, which may also be included in the factorization of a_i. In fact, gcd(a_i, b*a_j)=gcd(a_i/gcd(a_i, a_j), b)*gcd(a_i, a_j), I think. But I can't see a way to make use of this. :)
Anyhow, in our construction, m/a_i is simply a shortcut to calculate the product of all a_j, where j=1..1, j!=i. As a result, gcd(m/a_i, a_i) contains all gcd(a_i, a_j) as a factor. So, obviously, the maximum of these individual gcd results will divide g_i.
Now, the largest g_i is of particular interest to us: it is either the maximum gcd itself (if x_i is 1), or a good candidate for being one. To do that, we do another n-1 gcd operations, and calculate r_i explicitly. Then, we drop all g_j less than or equal to r_i as candidates. If we don't have any other candidate left, we are done. If not, we pick up the next largest g_k, and calculate r_k. If r_k <= r_i, we drop g_k, and repeat with another g_k'. If r_k > r_i, we filter out remaining g_j <= r_k, and repeat.
I think it is possible to construct a number set that will make this algorithm run in O(n^2) (if we fail to filter out anything), but on random number sets, I think it will quickly get rid of large chunks of candidates.
pseudocode
function getGcdMax(array[])
arrayUB=upperbound(array)
if (arrayUB<1)
error
pointerA=0
pointerB=1
gcdMax=0
do
gcdMax=MAX(gcdMax,gcd(array[pointera],array[pointerb]))
pointerB++
if (pointerB>arrayUB)
pointerA++
pointerB=pointerA+1
until (pointerB>arrayUB)
return gcdMax
I am given an array of real numbers, A. It has n+1 elements.
It is known that there are at least 2 elements of the array, x and y, such that:
abs(x-y) <= (max(A)-min(A))/n
I need to create an algorithm for finding the 2 items (if there are more, any couple is good) in O(n) time.
I've been trying for a few hours and I'm stuck, any clues/hints?
woo I got it! The trick is in the Pigeonhole Principle.
Okay.. think of the numbers as being points on a line. Then min(A) and max(A) define the start and end points of the line respectively. Now divide that line into n equal intervals of length (max(A)-min(A))/n. Since there are n+1 points, two of them must fall into one of the intervals.
Note that we don't need to rely on the question telling us that there are two points that satisfy the criterion. There are always two points that satisfy it.
The algorithm itself: You can use a simplified form of bucket sort here, since you only need one item per bucket (hit two and you're done). First loop once through the array to get min(A) and max(A) and create an integer array buckets[n] initialized to some default value, say -1. Then go for a second pass:
for (int i=0; i<len; i++) {
int bucket_num = find_bucket(array[i]);
if (bucket[bucket_num] == -1)
bucket[bucket_num] = i;
else
// found pair at (i, bucket[bucket_num])
}
Where find_bucket(x) returns the rounded-down integer result of x / ((max(A)-min(A))/n).
Let's re-word the problem: we're to find two elements, such that abs(x-y) <= c, where c is a constant, that we can find in O(n) time. (Indeed, we can compute both max(A) and min(A) in linear time and just assign c=(max-min)/n).
Let's imagine we have a set of buckets, so that in first bucket elements 0<=x<c are placed, in the second bucket elements c<=x<=2c are placed, etc. For each element, we can determine its bucket for O(1) time. Note that the number of buckets occupied will be not more than the number of elements in array.
Let's iterate the array and place each element to its bucket. If in the bucket we're going to place it, there already is another element, then we've just found the proper pair of x and y!
If we've iterated the whole array and every element has fallen into its own bucket, no worries! Iterate the buckets now (there is not more than n buckets, as we've said above) and for each bucket element x, if in the next bucket y element is such that abs(x-y)<=c, then we've found the solution.
If we iterated all the buckets and found no proper elements, then there is no solution. OMG, I really missed that pigeonhole stuff (see the other answer).
Buckets may be implemented as a hash map, where each bucket holds one array index (placing element in bucket will look like this: buckets[ a[i] / c] = i). We compute c in O(n) time, assign items to buckets in O(n)*O(1) time (O(1) is access to hash map), traverse buckets in O(n) time. Therefore, the whole algorithm is linear.