¿Cómo encontrar 5 valores repetidos en el tiempo O (n)?

15

Suponga que tiene una matriz de tamaño n 6 quen6 contiene enteros de a , inclusive, con exactamente cinco repetidos. Necesito proponer un algoritmo que pueda encontrar los números repetidos en el tiempo . No puedo, por mi vida, pensar en nada. Creo que ordenar, en el mejor de los casos, sería ? Luego, atravesar la matriz sería , lo que da como resultado . Sin embargo, no estoy muy seguro de si sería necesario ordenar, ya que he visto algunas cosas difíciles con listas vinculadas, colas, pilas, etc.11n -5n5O(n)O(n)O(nlogn)O(nlogn)O(n)O(n)O(n2logn)O(n2logn)

darylnak
fuente
16
O ( n log n ) + O ( n ) O ( n 2 log n ) O ( n log n ) O ( n 2 log n )O(nlogn)+O(n) no es . Es . Sería si hicieras el orden n veces. O(n2logn)O(nlogn)O(n2logn)
Financia la demanda de Mónica el
1
@leftaroundabout Estos algoritmos son O ( k n )O(kn) donde nn es el tamaño de la matriz ykk es el tamaño del conjunto de entrada. dado que k = n - c o n s t a n tk=nconstant estos algoritmos funcionan en O ( n 2 )O(n2)
Roman Gräf
44
@ RomanGräf parece que la situación real es esta: los algoritmos funcionan en O ( log k n )O(logkn) , donde kk es el tamaño del dominio. Entonces, para un problema como los OP, todo se reduce a lo mismo si usa un algoritmo de este tipo en el dominio de tamaño nn , o un algoritmo tradicional O ( n log n )O(nlogn) en un dominio de tamaño ilimitado. Tiene sentido también.
Leftaroundabout
55
Para n = 6 , el único número permitido es 1 , según su descripción. Pero entonces 1 tendría que repetirse seis, no cinco veces. n=611
Alex Reinking

Respuestas:

22

Puede crear una matriz B adicional de tamaño n . Inicialmente establezca todos los elementos de la matriz en 0 . Luego recorra la matriz de entrada A y aumente B [ A [ i ] ] en 1 para cada i . Después de eso, simplemente verifique la matriz B : repita sobre A y si B [ A [ i ] ] > 1, entonces se repite A [ i ] . Lo resuelves en O ( n )Bn0AB[A[i]]iBAB[A[i]]>1A[i]O(n)tiempo a costa de la memoria que es O ( n ) y porque sus enteros están entre 1 y n - 5 .O(n)1n5

fade2black
fuente
26

La solución en la respuesta de fade2black es la estándar, pero utiliza el espacio O ( n ) . Puede mejorar esto al espacio O ( 1 ) de la siguiente manera:O(n)O(1)

  1. Let the array be A[1],,A[n]A[1],,A[n]. For d=1,,5d=1,,5, compute σd=ni=1A[i]dσd=ni=1A[i]d.
  2. Compute τd=σdn5i=1idτd=σdn5i=1id (you can use the well-known formulas to compute the latter sum in O(1)O(1)). Note that τd=md1++md5τd=md1++md5, where m1,,m5m1,,m5 are the repeated numbers.
  3. Compute the polynomial P(t)=(tm1)(tm5)P(t)=(tm1)(tm5). The coefficients of this polynomial are symmetric functions of m1,,m5m1,,m5 which can be computed from τ1,,τ5τ1,,τ5 in O(1)O(1).
  4. Find all roots of the polynomial P(t)P(t) by trying all n5n5 possibilities.

This algorithm assumes the RAM machine model, in which basic arithmetic operations on O(logn)O(logn)-bit words take O(1)O(1) time.


Another way to formulate this solution is along the following lines:

  1. Calculate x1=ni=1A[i]x1=ni=1A[i], and deduce y1=m1++m5y1=m1++m5 using the formula y1=x1n5i=1iy1=x1n5i=1i.
  2. Calculate x2=1i<jA[i]A[j]x2=1i<jA[i]A[j] in O(n)O(n) using the formula x2=(A[1])A[2]+(A[1]+A[2])A[3]+(A[1]+A[2]+A[3])A[4]++(A[1]++A[n1])A[n].
    x2=(A[1])A[2]+(A[1]+A[2])A[3]+(A[1]+A[2]+A[3])A[4]++(A[1]++A[n1])A[n].
  3. Deduce y2=1i<j5mimjy2=1i<j5mimj using the formula y2=x21i<jn5ij(n5i=1i)y1.
    y2=x21i<jn5ij(i=1n5i)y1.
  4. Calculate x3,x4,x5x3,x4,x5 and deduce y3,y4,y5y3,y4,y5 along similar lines.
  5. The values of y1,,y5y1,,y5 are (up to sign) the coefficients of the polynomial P(t)P(t) from the preceding solution.

This solution shows that if we replace 5 by dd, then we get (I believe) a O(d2n)O(d2n) algorithm using O(d2)O(d2) space, which performs O(dn)O(dn) arithmetic operations on integers of bit-length O(dlogn)O(dlogn), keeping at most O(d)O(d) of these at any given time. (This requires careful analysis of the multiplications we perform, most of which involve one operand of length only O(logn)O(logn).) It is conceivable that this can be improved to O(dn)O(dn) time and O(d)O(d) space using modular arithmetic.

Yuval Filmus
fuente
Any interpretation of σdσd and τdτd, P(t)P(t), mimi and so on? Why d{1,2,3,4,5}d{1,2,3,4,5}?
styrofoam fly
3
The insight behind the solution is the summing trick, which appears in many exercises (for example, how do you find the missing element from an array of length n1n1 containing all but one of the numbers 1,,n1,,n?). The summing trick can be used to compute f(m1)++f(m5)f(m1)++f(m5) for an arbitrary function ff, and the question is which ff to choose in order to be able to deduce m1,,m5m1,,m5. My answer uses familiar tricks from the elementary theory of symmetric functions.
Yuval Filmus
1
@hoffmale Actually, O(d2)O(d2).
Yuval Filmus
1
@hoffmale Each of them takes dd machine words.
Yuval Filmus
1
@BurnsBA The problem with this approach is that (n5)#(n5)# is much larger than (n4)(n5)2(n4)(n5)2. Operations on large numbers are slower.
Yuval Filmus
8

There's also a linear time and constant space algorithm based on partitioning, which may be more flexible if you're trying to apply this to variants of the problem that the mathematical approach doesn't work well on. This requires mutating the underlying array and has worse constant factors than the mathematical approach. More specifically, I believe the costs in terms of the total number of values nn and the number of duplicates dd are O(nlogd)O(nlogd) and O(d)O(d) respectively, though proving it rigorously will take more time than I have at the moment.


Algorithm

Start with a list of pairs, where the first pair is the range over the whole array, or [(1,n)][(1,n)] if 1-indexed.

Repeat the following steps until the list is empty:

  1. Take and remove any pair (i,j)(i,j) from the list.
  2. Find the minimum and maximum, minmin and maxmax, of the denoted subarray.
  3. If min=maxmin=max, the subarray consists only of equal elements. Yield its elements except one and skip steps 4 to 6.
  4. If maxmin=jimaxmin=ji, the subarray contains no duplicates. Skip steps 5 and 6.
  5. Partition the subarray around min+max2min+max2, such that elements up to some index kk are smaller than the separator and elements above that index are not.
  6. Add (i,k)(i,k) and (k+1,j)(k+1,j) to the list.

Cursory analysis of time complexity.

Steps 1 to 6 take O(ji)O(ji) time, since finding the minimum and maximum and partitioning can be done in linear time.

Every pair (i,j)(i,j) in the list is either the first pair, (1,n)(1,n), or a child of some pair for which the corresponding subarray contains a duplicate element. There are at most dlog2n+1dlog2n+1 such parents, since each traversal halves the range in which a duplicate can be, so there are at most 2dlog2n+12dlog2n+1 total when including pairs over subarrays with no duplicates. At any one time, the size of the list is no more than 2d2d.

Consider the work to find any one duplicate. This consists of a sequence of pairs over an exponentially decreasing range, so the total work is the sum of the geometric sequence, or O(n)O(n). This produces an obvious corollary that the total work for dd duplicates must be O(nd)O(nd), which is linear in nn.

To find a tighter bound, consider the worst-case scenario of maximally spread out duplicates. Intuitively, the search takes two phases, one where the full array is being traversed each time, in progressively smaller parts, and one where the parts are smaller than ndnd so only parts of the array are traversed. The first phase can only be logdlogd deep, so has cost O(nlogd)O(nlogd), and the second phase has cost O(n)O(n) because the total area being searched is again exponentially decreasing.

Veedrac
fuente
Thank you for the explanation. Now I understand. A very pretty algorithm!
D.W.
5

Leaving this as an answer because it needs more space than a comment gives.

You make a mistake in the OP when you suggest a method. Sorting a list and then transversing it O(nlogn)O(nlogn) time, not O(n2logn)O(n2logn) time. When you do two things (that take O(f)O(f) and O(g)O(g) respectively) sequentially then the resulting time complexity is O(f+g)=O(maxf,g)O(f+g)=O(maxf,g) (under most circumstances).

In order to multiply the time complexities, you need to be using a for loop. If you have a loop of length ff and for each value in the loop you do a function that takes O(g)O(g), then you'll get O(fg)O(fg) time.

So, in your case you sort in O(nlogn)O(nlogn) and then transverse in O(n)O(n) resulting in O(nlogn+n)=O(nlogn)O(nlogn+n)=O(nlogn). If for each comparison of the sorting algorithm you had to do a computation that takes O(n)O(n), then it would take O(n2logn)O(n2logn) but that's not the case here.


In case your curious about my claim that O(f+g)=O(maxf,g)O(f+g)=O(maxf,g), it's important to note that that's not always true. But if fO(g)fO(g) or gO(f)gO(f) (which holds for a whole host of common functions), it will hold. The most common time it doesn't hold is when additional parameters get involved and you get expressions like O(2cn+nlogn)O(2cn+nlogn).

Stella Biderman
fuente
3

There's an obvious in-place variant of the boolean array technique using the order of the elements as the store (where arr[x] == x for "found" elements). Unlike the partition variant that can be justified for being more general I'm unsure when you'd actually need something like this, but it is simple.

for idx from n-4 to n
    while arr[arr[idx]] != arr[idx]
        swap(arr[arr[idx]], arr[idx])

This just repeatedly puts arr[idx] at the location arr[idx] until you find that location already taken, at which point it must be a duplicate. Note that the total number of swaps is bounded by nn since each swap makes its exit condition correct.

Veedrac
fuente
You're going to have to give some sort of argument that the inner while loop runs in constant time on average. Otherwise, this isn't a linear-time algorithm.
David Richerby
@DavidRicherby It doesn't run constant time on average, but the outer loop only runs 5 times so that's fine. Note that the total number of swaps is bounded by nn since each swap makes its exit condition correct, so even if the number of duplicate values increases the total time is still linear (aka. it takes nn steps rather than ndnd) .
Veedrac
Oops, I somehow didn't notice that the outer loop runs a constant number of times! (Edited to include your note about the number of swaps and also so I could reverse my downvote.)
David Richerby
1

Subtract the values you have from the sum ni=1i=(n1)n2ni=1i=(n1)n2.

So, after Θ(n)Θ(n) time (assuming arithmetic is O(1), which it isn't really, but let's pretend) you have a sum σ1σ1 of 5 integers between 1 and n:

x1+x2+x3+x4+x5=σ1x1+x2+x3+x4+x5=σ1

Supposedly, this is no good, right? You can't possibly figure out how to break this up into 5 distinct numbers.

Ah, but this is where it gets to be fun! Now do the same thing as before, but subtract the squares of the values from ni=1i2ni=1i2. Now you have:

x12+x22+x32+x42+x52=σ2x12+x22+x32+x42+x52=σ2

See where I'm going with this? Do the same for powers 3, 4 and 5 and you have yourself 5 independent equations in 5 variables. I'm pretty sure you can solve for xx⃗ .

Caveats: Arithmetic is not really O(1). Also, you need a bit of space to represent your sums; but not as much as you would imagine - you can do most everything modularly, as long as you have, oh, log(5n6)log(5n6) bits; that should do it.

einpoklum
fuente
Doesn't @YuvalFilmus propose the same solution?
fade2black
@fade2black: Oh, yes, it does, sorry, I just saw the first line of his solution.
einpoklum
0

Easiest way to solve the problem is to create array in which we will count the apperances for each number in the original array, and then traverse all number from 11 to n5n5 and check if the number appears more than once, the complexity for this solution in both memory and time is linear, or O(N)O(N)

someone12321
fuente
1
This is the same @fade2black's answer (although a bit easier on the eyes)
LangeHaare
0

Map an array to 1 << A[i] and then XOR everything together. Your duplicates will be the numbers where corresponding bit is off.

Hauleth
fuente
There are five duplicates, so the xor trick will not break in some cases.
Evil
1
The running time of this is O(n2)O(n2). Each bitvector is nn bits long, so you each bitvector operation takes O(n)O(n) time, and you do one bit vector operation per element of the original array, for a total of O(n2)O(n2) time.
D.W.
@D.W. But given that the machines we normally use are fixed at either 32 or 64-bits, and these don't change at run-time (i.e. they're constant), why shouldn't they be treated as such and assume that the bit operations are in O(1)O(1) instead of O(n)O(n)?
code_dredd
1
@ray, I think you answered your own question. Given that the machines we normally use are fixed at 64-bits, the running time to do an operation on a nn-bit vector is O(n), not O(1). It takes something like n/64 instructions to do some operation on all n bits of a n-bit vector, and n/64 is O(n), not O(1).
D.W.
@D.W. What I got out of prev. comments was that a bit vector referred to a single element in an n-sized array, with the bit vector being 64-bits, which would be the constant I'm referring to. Obviously, processing an an array of size n will take O(kn) time, if we assume there're k-bits per element and n the number of elements in the array. But k=64, so an operation for an array element w/ a constant bit count should be O(1) instead of O(k) and the array O(n) instead of O(kn). Are you keeping the k for the sake of completeness/correctness or am I missing something else?
code_dredd
-2
DATA=[1,2,2,2,2,2]

from collections import defaultdict

collated=defaultdict(list):
for item in DATA:
    collated[item].append(item)
    if len(collated) == 5:
        return item.

# n time
user78484
fuente
4
Welcome to the site. We're a computer science site, so we're looking for algorithms and explanations, not code dumps that require understanding of a particular language and its libraries. In particular, your claim that this code runs in linear time assumes that collated[item].append(item) runs in constant time. Is that really true?
David Richerby
3
Also, you are looking for a value which is repeated five times. In contrast, the OP is looking for five values, which are each repeated twice.
Yuval Filmus