I've been trying to solve a problem in combinations. I have a matrix 6X6 i'm trying to find all combinations of length 8 in the matrix.
I have to move from neighbor to neighbor form each row,column position and i wrote a recursive program which generates the combination but the problem is it generates a lot of duplicates as well and hence is inefficient. I would like to know how could i eliminate calculating duplicates and save time.
int a={{1,2,3,4,5,6},
{8,9,1,2,3,4},
{5,6,7,8,9,1},
{2,3,4,5,6,7},
{8,9,1,2,3,4},
{5,6,7,8,9,1},
}
void genSeq(int row,int col,int length,int combi)
{
if(length==8)
{
printf("%d\n",combi);
return;
}
combi = (combi * 10) + a[row][col];
if((row-1)>=0)
genSeq(row-1,col,length+1,combi);
if((col-1)>=0)
genSeq(row,col-1,length+1,combi);
if((row+1)<6)
genSeq(row+1,col,length+1,combi);
if((col+1)<6)
genSeq(row,col+1,length+1,combi);
if((row+1)<6&&(col+1)<6)
genSeq(row+1,col+1,length+1,combi);
if((row-1)>=0&&(col+1)<6)
genSeq(row-1,col+1,length+1,combi);
if((row+1)<6&&(row-1)>=0)
genSeq(row+1,col-1,length+1,combi);
if((row-1)>=0&&(col-1)>=0)
genSeq(row-1,col-1,length+1,combi);
}
I was also thinking of writing a dynamic program basically recursion with memorization. Is it a better choice?? if yes than I'm not clear how to implement it in recursion. Have i really hit a dead end with approach???
Thankyou
Edit
Eg result
12121212,12121218,12121219,12121211,12121213.
the restrictions are that you have to move to your neighbor from any point, you have to start for each position in the matrix i.e each row,col. you can move one step at a time, i.e right, left, up, down and the both diagonal positions. Check the if conditions.
i.e
if your in (0,0) you can move to either (1,0) or (1,1) or (0,1) i.e three neighbors.
if your in (2,2) you can move to eight neighbors.
so on...
To eliminate duplicates you can covert 8 digit sequences into 8-digit integers and put them in a hashtable.
Memoization might be a good idea. You can memoize for each cell in the matrix all possible combinations of length 2-7 that can be achieved from it. Going backwards: first generate for each cell all sequences of 2 digits. Then based on that of 3 digits etc.
UPDATE: code in Python
# original matrix
lst = [
[1,2,3,4,5,6],
[8,9,1,2,3,4],
[5,6,7,8,9,1],
[2,3,4,5,6,7],
[8,9,1,2,3,4],
[5,6,7,8,9,1]]
# working matrtix; wrk[i][j] contains a set of all possible paths of length k which can end in lst[i][j]
wrk = [[set() for i in range(6)] for j in range(6)]
# for the first (0rh) iteration initialize with single step paths
for i in range(0, 6):
for j in range(0, 6):
wrk[i][j].add(lst[i][j])
# run iterations 1 through 7
for k in range(1,8):
# create new emtpy wrk matrix for the next iteration
nw = [[set() for i in range(6)] for j in range(6)]
for i in range(0, 6):
for j in range(0, 6):
# the next gen. wrk[i][j] is going to be based on the current wrk paths of its neighbors
ns = set()
if i > 0:
for p in wrk[i-1][j]:
ns.add(10**k * lst[i][j] + p)
if i < 5:
for p in wrk[i+1][j]:
ns.add(10**k * lst[i][j] + p)
if j > 0:
for p in wrk[i][j-1]:
ns.add(10**k * lst[i][j] + p)
if j < 5:
for p in wrk[i][j+1]:
ns.add(10**k * lst[i][j] + p)
nw[i][j] = ns
wrk = nw
# now build final set to eliminate duplicates
result = set()
for i in range(0, 6):
for j in range(0, 6):
result |= wrk[i][j]
print len(result)
print result
There are LOTS of ways to do this. Going through every combination is a perfectly reasonable first approach. It all depends on your requirements. If your matrix is small, and this operation isn't time sensitive, then there's no problem.
I'm not really an algorithms guy, but I'm sure there are really clever ways of doing this that someone will post after me.
Also, in Java when using CamelCase, method names should start with a lowercase character.
int a={{1,2,3,4,5,6},
{8,9,1,2,3,4},
{5,6,7,8,9,1},
{2,3,4,5,6,7},
{8,9,1,2,3,4},
{5,6,7,8,9,1},
}
By length you mean summation of combination of matrix elements resulting 8. i.e., elements to sum up 8 with in row itself and with the other row elements. From row 1 = { {2,6}, {3,5}, } and now row 1 elements with row 2 and so on. Is that what you are expecting ?
You can think about your matrix like it is one-dimension array - no matter here ("place" the rows one by one). For one-dimension array you can write a function like (assuming you should print the combinations)
f(i, n) prints all combinations of length n using elements a[i] ... a[last].
It should skip some elements from a[i] to a[i + k] (for all possible k), print a[k] and make a recursive call f(i + k + 1, n - 1).
Related
You are given a permutation p1,p2,...,pn of numbers from 1 to n.
A permutation is a sequence of integers from 1 to n of length n containing each number exactly once.
You are given q queries where each query consists of two integers a and b, In response to each query you need to return a number of inversions of permutation after swapping elements at index a and b, Here every query is independent i.e. after each query the permutation is restored to its initial state.
An inversion in a permutation p is a pair of indices (i, j) such that i > j and pi < pj. For example, a permutation [4, 1, 3, 2] contains 4 inversions: (2, 1), (3, 1), (4, 1), (4, 3).
Input: The first line contains n,q.
The second line contains the space-separated permutation p1,p2,...,pn.
Each line of the next q lines contains two integers a,b.
Output: For each query Print an integer denoting the number of Inversion on a new line.
Sample input:
5 5
1 2 3 4 5
1 2
1 3
2 5
2 4
3 3
Output:
1
3
5
3
0
Constraints:
2<=n<=1000
1<=q<=200000
My approach: I am counting no of inversions using BIT (https://www.geeksforgeeks.org/count-inversions-array-set-3-using-bit/) for each query after swapping elements at position a and b..and then again swapping it so that my array remains unchanged. But this solution gives TLE for large test cases. Is there any better approach for this problem?
You are getting TLE probably because number of computations in this approach is q * (n * log(n)) = 2 * 10^5 * 10^3 * log(1000) = ~10^9, which is more than generally accepted computations ~10^8.
I can think of the following solution. Please note that I have not coded / verified it:
Denoting ri == number of indices j, such that i > j && pi < pj. Eg: [2, 3, 1, 4], r3 = 2. Basically, it means the number of inversions with the farther index as i. (Please note that I am using 1-based index as per the question. Also,a < b as per the question)
Thus we have: Sum of ri == #invs (number of inversions)
We can calculate initial total #invs in O(n^2)
When a and b are swapped, we can observe that:
a) ri remains constant, where i < a .
b) ri remains constant, where i > b.
Only ri changes where a <= i <=b, and that too on these following conditions. I am considering the case when pa < pb. Exact opposite case will need to considered when pa > pb.
a) Since pa < pb, thus this swap causes #invs = #invs + 1
b) If (pi < pa && pi < pb) || (pi > pa && pi > pb), this swap does not change ri. Eg: [2,....10,....5]. Here Swapping 2 and 5 does not change the r value for 10.
c) If pa < pi < pb, it will increment ri by 1, and new rb by 1. Eg: [2,....3,.....4], when 2 and 4 are swapped, we have [4,....3,....2], the rvalue 3 increases by 1 (because of 4); and also the r value of 2 increase by 1 (because of 3). Please note that increment because of what about 4 > 2? was already calculated in step (a), and needs to be done once only.
d) We need to find all such indicies i where pa < pi < pb as we started with above. Let us call it f(a, b). Then the total change in #invs delta = (2 * f(a, b)) + 1, and answer will be #original_invs + delta.
As I mentioned, all the exact opposite steps need to be done for the case pa > pb. The delta will be negative in that case.
Now, the only thing remained is to solve: Given a, b, find f(a, b) efficiently. For this, we can pre-process and store it for all pairs of indices. This will take O(N^2) space, and O(N^2 * log(N)) time, using a balanced binary-search-tree (BST). Again showing steps for pre-processing for case pa < pb only. Another set of pre-processing steps needs to be done for the other case:
We will use self-balancing BST, in which each node also contains the following fields:
a) field_1: This denotes the size of the left sub-tree. This value will be updated on every insert operation, if size of left-sub-tree changes.
b) field_2: This denotes the number of elements < node.value that this tree has. This value is initialized once when the node is inserted and does not change thereafter. I have added a small explanation of how it will be achieved in Addendum-A. This field is basically our pre-processing, which will determine f(a, b).
With all of this now, for each index i, where 0 <= i < n, do the following: Create new tree. Insert pj values into the tree one by one, where (i < j < n ) && (pa < pj) . (Please note we are not inserting values where pa > pj). The method given in Addendum-A will make sure we find f(i, j) while inserting.
There will be n such pre-processed trees, one for every index. For finding f(a, b): We need to look into ath tree, and search node.value = pb. This node's field_2 = f(a, b).
The complexity of insertion is O(logN). So, the total pre-processing computation = O(N * N(logN)). Search is O(logN), so the query complexity is O(q * logN). Total complexity = O(N^2) + O(N * N (logN)) + O(q * logN) which will turn out ~10^7
==============================================================================
Addendum A: How to populate field_2 while inserting node:
i) Insert the node, and balance the tree. Update field_1 as required.
i) Initailze ans = 0. Traverse the BST from root searching for your node.
iii) do {
If node.value < search_key_b, ans += node.left_subtree_size + 1
} while(!node.found)
iv) ans -= 1
We can solve this in O(n log n) space and O(n log n + Q * log^2(n)) time with a merge-sort tree. The merge-sort tree allows us to find the number of elements inside a subarray that are greater than or lower than an input number in O(log^2(n)) time and O(n log n) space.
First we record the total number of inversions in O(n log n) time, for which there are known methods. To query the effect of a swap bound by left and right, consider the subarray between:
subtract the number of elements greater
than right in the subarray (those will
no longer be inversions)
subtract the number of elements smaller
than left in the subarray (those will
no longer be inversions)
add the number of elements greater
than left in the subarray (those will
be new inversions)
add the number of elements smaller
than right in the subarray (those will
be new inversions)
if right > left, add 1
if left > right, subtract 1
I am looking for a least time-complex algorithm that would solve a variant of the perfect sum problem (initially: finding all variable size subset combinations from an array [*] of integers of size n that sum to a specific number x) where the subset combination size is of a fixed size k and return the possible combinations without direct and also indirect (when there's a combination containing the exact same elements from another in another order) duplicates.
I'm aware this problem is NP-hard, so I am not expecting a perfect general solution but something that could at least run in a reasonable time in my case, with n close to 1000 and k around 10
Things I have tried so far:
Finding a combination, then doing successive modifications on it and its modifications
Let's assume I have an array such as:
s = [1,2,3,3,4,5,6,9]
So I have n = 8, and I'd like x = 10 for k = 3
I found thanks to some obscure method (bruteforce?) a subset [3,3,4]
From this subset I'm finding other possible combinations by taking two elements out of it and replacing them with other elements that sum the same, i.e. (3, 3) can be replaced by (1, 5) since both got the same sum and the replacing numbers are not already in use. So I obtain another subset [1,5,4], then I repeat the process for all the obtained subsets... indefinitely?
The main issue as suggested here is that it's hard to determine when it's done and this method is rather chaotic. I imagined some variants of this method but they really are work in progress
Iterating through the set to list all k long combinations that sum to x
Pretty self explanatory. This is a naive method that do not work well in my case since I have a pretty large n and a k that is not small enough to avoid a catastrophically big number of combinations (the magnitude of the number of combinations is 10^27!)
I experimented several mechanism related to setting an area of research instead of stupidly iterating through all possibilities, but it's rather complicated and still work in progress
What would you suggest? (Snippets can be in any language, but I prefer C++)
[*] To clear the doubt about whether or not the base collection can contain duplicates, I used the term "array" instead of "set" to be more precise. The collection can contain duplicate integers in my case and quite much, with 70 different integers for 1000 elements (counts rounded), for example
With reasonable sum limit this problem might be solved using extension of dynamic programming approach for subset sum problem or coin change problem with predetermined number of coins. Note that we can count all variants in pseudopolynomial time O(x*n), but output size might grow exponentially, so generation of all variants might be a problem.
Make 3d array, list or vector with outer dimension x-1 for example: A[][][]. Every element A[p] of this list contains list of possible subsets with sum p.
We can walk through all elements (call current element item) of initial "set" (I noticed repeating elements in your example, so it is not true set).
Now scan A[] list from the last entry to the beginning. (This trick helps to avoid repeating usage of the same item).
If A[i - item] contains subsets with size < k, we can add all these subsets to A[i] appending item.
After full scan A[x] will contain subsets of size k and less, having sum x, and we can filter only those of size k
Example of output of my quick-made Delphi program for the next data:
Lst := [1,2,3,3,4,5,6,7];
k := 3;
sum := 10;
3 3 4
2 3 5 //distinct 3's
2 3 5
1 4 5
1 3 6
1 3 6 //distinct 3's
1 2 7
To exclude variants with distinct repeated elements (if needed), we can use non-first occurence only for subsets already containing the first occurence of item (so 3 3 4 will be valid while the second 2 3 5 won't be generated)
I literally translate my Delphi code into C++ (weird, I think :)
int main()
{
vector<vector<vector<int>>> A;
vector<int> Lst = { 1, 2, 3, 3, 4, 5, 6, 7 };
int k = 3;
int sum = 10;
A.push_back({ {0} }); //fictive array to make non-empty variant
for (int i = 0; i < sum; i++)
A.push_back({{}});
for (int item : Lst) {
for (int i = sum; i >= item; i--) {
for (int j = 0; j < A[i - item].size(); j++)
if (A[i - item][j].size() < k + 1 &&
A[i - item][j].size() > 0) {
vector<int> t = A[i - item][j];
t.push_back(item);
A[i].push_back(t); //add new variant including current item
}
}
}
//output needed variants
for (int i = 0; i < A[sum].size(); i++)
if (A[sum][i].size() == k + 1) {
for (int j = 1; j < A[sum][i].size(); j++) //excluding fictive 0
cout << A[sum][i][j] << " ";
cout << endl;
}
}
Here is a complete solution in Python. Translation to C++ is left to the reader.
Like the usual subset sum, generation of the doubly linked summary of the solutions is pseudo-polynomial. It is O(count_values * distinct_sums * depths_of_sums). However actually iterating through them can be exponential. But using generators the way I did avoids using a lot of memory to generate that list, even if it can take a long time to run.
from collections import namedtuple
# This is a doubly linked list.
# (value, tail) will be one group of solutions. (next_answer) is another.
SumPath = namedtuple('SumPath', 'value tail next_answer')
def fixed_sum_paths (array, target, count):
# First find counts of values to handle duplications.
value_repeats = {}
for value in array:
if value in value_repeats:
value_repeats[value] += 1
else:
value_repeats[value] = 1
# paths[depth][x] will be all subsets of size depth that sum to x.
paths = [{} for i in range(count+1)]
# First we add the empty set.
paths[0][0] = SumPath(value=None, tail=None, next_answer=None)
# Now we start adding values to it.
for value, repeats in value_repeats.items():
# Reversed depth avoids seeing paths we will find using this value.
for depth in reversed(range(len(paths))):
for result, path in paths[depth].items():
for i in range(1, repeats+1):
if count < i + depth:
# Do not fill in too deep.
break
result += value
if result in paths[depth+i]:
path = SumPath(
value=value,
tail=path,
next_answer=paths[depth+i][result]
)
else:
path = SumPath(
value=value,
tail=path,
next_answer=None
)
paths[depth+i][result] = path
# Subtle bug fix, a path for value, value
# should not lead to value, other_value because
# we already inserted that first.
path = SumPath(
value=value,
tail=path.tail,
next_answer=None
)
return paths[count][target]
def path_iter(paths):
if paths.value is None:
# We are the tail
yield []
else:
while paths is not None:
value = paths.value
for answer in path_iter(paths.tail):
answer.append(value)
yield answer
paths = paths.next_answer
def fixed_sums (array, target, count):
paths = fixed_sum_paths(array, target, count)
return path_iter(paths)
for path in fixed_sums([1,2,3,3,4,5,6,9], 10, 3):
print(path)
Incidentally for your example, here are the solutions:
[1, 3, 6]
[1, 4, 5]
[2, 3, 5]
[3, 3, 4]
You should first sort the so called array. Secondly, you should determine if the problem is actually solvable, to save time... So what you do is you take the last k elements and see if the sum of those is larger or equal to the x value, if it is smaller, you are done it is not possible to do something like that.... If it is actually equal yes you are also done there is no other permutations.... O(n) feels nice doesn't it?? If it is larger, than you got a lot of work to do..... You need to store all the permutations in an seperate array.... Then you go ahead and replace the smallest of the k numbers with the smallest element in the array.... If this is still larger than x then you do it for the second and third and so on until you get something smaller than x. Once you reach a point where you have the sum smaller than x, you can go ahead and start to increase the value of the last position you stopped at until you hit x.... Once you hit x that is your combination.... Then you can go ahead and get the previous element so if you had 1,1,5, 6 in your thingy, you can go ahead and grab the 1 as well, add it to your smallest element, 5 to get 6, next you check, can you write this number 6 as a combination of two values, you stop once you hit the value.... Then you can repeat for the others as well.... You problem can be solved in O(n!) time in the worst case.... I would not suggest that you 10^27 combinations, meaning you have more than 10^27 elements, mhmmm bad idea do you even have that much space??? That's like 3bits for the header and 8 bits for each integer you would need 9.8765*10^25 terabytes just to store that clossal array, more memory than a supercomputer, you should worry about whether your computer can even store this monster rather than if you can solve the problem, that many combinations even if you find a quadratic solution it would crash your computer, and you know what quadratic is a long way off from O(n!)...
A brute force method using recursion might look like this...
For example, given variables set, x, k, the following pseudo code might work:
setSumStructure find(int[] set, int x, int k, int setIdx)
{
int sz = set.length - setIdx;
if (sz < x) return null;
if (sz == x) check sum of set[setIdx] -> set[set.size] == k. if it does, return the set together with the sum, else return null;
for (int i = setIdx; i < set.size - (k - 1); i++)
filter(find (set, x - set[i], k - 1, i + 1));
return filteredSets;
}
Problem statement:
Input:
First two inputs are integers n and m. n is the number of knights fighting in the tournament (2 <= n <= 100000, 1 <= m <= n-1). m is the number of battles that will take place.
The next line contains n power levels.
The next m lines contain two integers l and r, indicating the range of knight positions to compete in the ith battle.
After each battle, all nights apart from the one with the highest power level will be eliminated.
The range for each battle is given in terms of the new positions of the knights, not the original positions.
Output:
Output m lines, the ith line containing the original positions (indices) of the knights from that battle. Each line is in ascending order.
Sample Input:
8 4
1 0 5 6 2 3 7 4
1 3
2 4
1 3
0 1
Sample Output:
1 2
4 5
3 7
0
Here is a visualisation of this process.
1 2
[(1,0),(0,1),(5,2),(6,3),(2,4),(3,5),(7,6),(4,7)]
-----------------
4 5
[(1,0),(6,3),(2,4),(3,5),(7,6),(4,7)]
-----------------
3 7
[(1,0),(6,3),(7,6),(4,7)]
-----------------
0
[(1,0),(7,6)]
-----------
[(7,6)]
I have solved this problem. My program produces the correct output, however, it is O(n*m) = O(n^2). I believe that if I erase knights more efficiently from the vector, efficiency can be increased. Would it be more efficient to erase elements using a set? I.e. erase contiguous segments rather that individual knights. Is there an alternative way to do this that is more efficient?
#define INPUT1(x) scanf("%d", &x)
#define INPUT2(x, y) scanf("%d%d", &x, &y)
#define OUTPUT1(x) printf("%d\n", x);
int main(int argc, char const *argv[]) {
int n, m;
INPUT2(n, m);
vector< pair<int,int> > knights(n);
for (int i = 0; i < n; i++) {
int power;
INPUT(power);
knights[i] = make_pair(power, i);
}
while(m--) {
int l, r;
INPUT2(l, r);
int max_in_range = knights[l].first;
for (int i = l+1; i <= r; i++) if (knights[i].first > max_in_range) {
max_in_range = knights[i].first;
}
int offset = l;
int range = r-l+1;
while (range--) {
if (knights[offset].first != max_in_range) {
OUTPUT1(knights[offset].second));
knights.erase(knights.begin()+offset);
}
else offset++;
}
printf("\n");
}
}
Well, removing from vector wouldn't be efficient for sure. Removing from set, or unordered set would be more effective (use iterators instead of indexes).
Yet the problem will still remain O(n^2), because you have two nested whiles running n*m times.
--EDIT--
I believe I understand the question now :)
First let's calculate the complexity of your code above. Your worst case would be the case that max range in all battles is 1 (two nights for each battle) and the battles are not ordered with respect to the position. Which means you have m battles (in this case m = n-1 ~= O(n))
The first while loop runs n times
For runs for once every time which makes it n*1 = n in total
The second while loop runs once every time which makes it n again.
Deleting from vector means n-1 shifts that makes it O(n).
Thus with the complexity of the vector total complexity is O(n^2)
First of all, you don't really need the inner for loop. Take the first knight as the max in range, compare the rest in the range one-by-one and remove the defeated ones.
Now, i believe it can be done in O(nlogn) with using std::map. The key to the map is the position and the value is the level of the knight.
Before proceeding, finding and removing an element in map is logarithmic, iterating is constant.
Finally, your code should look like:
while(m--) // n times
strongest = map.find(first_position); // find is log(n) --> n*log(n)
for (opponent = next of strongest; // this will run 1 times, since every range is 1
opponent in range;
opponent = next opponent) // iterating is constant
// removing from map is log(n) --> n * 1 * log(n)
if strongest < opponent
remove strongest, opponent is the new strongest
else
remove opponent, (be careful to remove it after iterating to next)
Ok, now the upper bound would be O(2*nlogn) = O(nlogn). If the ranges increases, that makes the run time of upper loop decrease but increases the number of remove operations. I'm sure the upper bound won't change, let's make it a homework for you to calculate :)
A solution with a treap is pretty straightforward.
For each query, you need to split the treap by implicit key to obtain the subtree that corresponds to the [l, r] range (it takes O(log n) time).
After that, you can iterate over the subtree and find the knight with the maximum strength. After that, you just need to merge the [0, l) and [r + 1, end) parts of the treap with the node that corresponds to this knight.
It's clear that all parts of the solution except for the subtree traversal and printing work in O(log n) time per query. However, each operation reinserts only one knight and erase the rest from the range, so the size of the output (and the sum of sizes of subtrees) is linear in n. So the total time complexity is O(n log n).
I don't think you can solve with standard stl containers because there'no standard container that supports getting an iterator by index quickly and removing arbitrary elements.
I am trying to understand the algorithm that gives me the number of increasing subsequences of length K in an array in time O(nklog(n)). I know how to solve this very same problem using the O(k*n^2) algorithm. I have looked up and found out this solution uses BIT (Fenwick Tree) and DP. I have also found some code, but I have not been able to understand it.
Here are some links I've visited that have been helpful.
Here in SO
Topcoder forum
Random webpage
I would really appreciate if some can help me out understand this algorithm.
I am reproducing my algorithm from here, where its logic is explained:
dp[i, j] = same as before num[i] = how many subsequences that end with i (element, not index this time)
have a certain length
for i = 1 to n do dp[i, 1] = 1
for p = 2 to k do // for each length this time num = {0}
for i = 2 to n do
// note: dp[1, p > 1] = 0
// how many that end with the previous element
// have length p - 1
num[ array[i - 1] ] += dp[i - 1, p - 1] *1*
// append the current element to all those smaller than it
// that end an increasing subsequence of length p - 1,
// creating an increasing subsequence of length p
for j = 1 to array[i] - 1 do *2*
dp[i, p] += num[j]
You can optimize *1* and *2* by using segment trees or binary indexed trees. These will be used to efficiently process the following operations on the num array:
Given (x, v) add v to num[x] (relevant for *1*);
Given x, find the sum num[1] + num[2] + ... + num[x] (relevant for *2*).
These are trivial problems for both data structures.
Note: This will have complexity O(n*k*log S), where S is the upper bound on the values in your array. This may or may not be good enough. To make it O(n*k*log n), you need to normalize the values of your array prior to running the above algorithm. Normalization means converting all of your array values into values lower than or equal to n. So this:
5235 223 1000 40 40
Becomes:
4 2 3 1 1
This can be accomplished with a sort (keep the original indexes).
Given a simple undirected graph like this:
Starting in D, A, B or C (V_start)—I have to calculate how many possible paths there are from the starting point (V_start) to the starting point (V_start) of n steps, where each edge and vertex can be visited an unlimited amount of times.
I was thinking of doing a depth first search, stopping when steps > n || (steps == n && vertex != V_start), however, this becomes rather expensive if, for instance, n = 1000000. My next thought led me to combining DFS with dynamic programming, however, this is where I'm stuck.
(This is not homework, just me getting stuck playing around with graphs and algorithms for the sake of learning.)
How would I go about solving this in a reasonable time with a large n?
This task is solved by matrix multiplication.
Create matrix nxn containing 0s and 1s (1 for a cell mat[i][j] if there is path from i to j). Multiply this matrix by itself k times (consider using fast matrix exponentiation). Then in the matrix's cell mat[i][j] you have the number of paths with length k starting from i and ending in j.
NOTE: The fast matrix exponentiation is basically the same as the fast exponentiation, just that instead you multiply numbers you multiply matrices.
NOTE2: Lets assume n is the number of vertices in the graph. Then the algorithm I propose here runs in time complexity O(log k * n3) and has memory complexity of O(n 2). You can improve it a bit more if you use optimized matrix multiplication as described here. Then the time complexity will become O(log k * nlog27).
EDIT As requested by Antoine I include an explanation why this algorithm actually works:
I will prove the algorithm by induction. The base of the induction is obvious: initially I have in the matrix the number of paths of length 1.
Let us assume that until the power of k if I raise the matrix to the power of k I have in mat[i][j] the number of paths with length k between i and j.
Now lets consider the next step k + 1. It is obvious that every path of length k + 1 consists of prefix of length k and one more edge. This basically means that the paths of length k + 1 can be calculated by (here I denote by mat_pow_k the matrix raised to the kth power)
num_paths(x, y, k + 1) = sumi=0i<n mat_pow_k[x][i] * mat[i][y]
Again: n is the number of vertices in the graph. This might take a while to understand, but basically the initial matrix has 1 in its mat[i][y] cell only if there is direct edge between x and y. And we count all possible prefixes of such edge to form path of length k + 1.
However the last thing I wrote is actually calculating the k + 1st power of mat, which proves the step of the induction and my statement.
It's quite like a dynamic programming problem:
define a f[n][m] to be the number of paths from the starting point to point n in m steps
from every point n to its adjacent k, you have formula: f[k][m+1] = f[k][m+1] + f[n][m]
in the initialization, all f[n][m] will be 0, but f[starting_point][0] = 1
so you can calculate the final result
pseudo code:
memset(f, 0, sizeof(f));
f[starting_point][0] = 1;
for (int step = 0; step < n; ++step) {
for (int point = 0; point < point_num; ++point) {
for (int next_point = 0; next_point < point_num; ++ next_point) {
if (adjacent[point][next_point]) {
f[next_point][step+1] += f[point][step];
}
}
}
}
return f[starting_point][n]