Background:
This is extra credit in a logic and algorithms class, we are currently covering propositional logic, P implies Q that kind of thing, so I think the Prof wanted to give us and assignment out of our depth.
I will implement this in C++, but right now I just want to understand whats going on in the example....which I don't.
Example
Enclosed is a walkthrough for the Lefty algorithm which computes the number
of nxn 0-1 matrices with t ones in each row and column, but none on the main
diagonal.
The algorithm used to verify the equations presented counts all the possible
matrices, but does not construct them.
It is called "Lefty", it is reasonably simple, and is best described with an
example.
Suppose we wanted to compute the number of 6x6 0-1 matrices with 2 ones
in each row and column, but no ones on the main diagonal. We first create a
state vector of length 6, filled with 2s:
(2 2 2 2 2 2)
This state vector symbolizes the number of ones we must yet place in each
column. We accompany it with an integer which we call the "puck", which is
initialized to 1. This puck will increase by one each time we perform a ones
placement in a row of the matrix (a "round"), and we will think of the puck as
"covering up" the column that we wonít be able to place ones in for that round.
Since we are starting with the first row (and hence the first round), we place
two ones in any column, but since the puck is 1, we cannot place ones in the
first column. This corresponds to the forced zero that we must place in the first
column, since the 1,1 entry is part of the matrixís main diagonal.
The algorithm will iterate over all possible choices, but to show each round,
we shall make a choice, say the 2nd and 6th columns. We then drop the state
vector by subtracting 1 from the 2nd and 6th values, and advance the puck:
(2 1 2 2 2 1); 2
For the second round, the puck is 2, so we cannot place a one in that column.
We choose to place ones in the 4th and 6th columns instead and advance the
puck:
(2 1 2 1 2 0); 3
Now at this point, we can place two ones anywhere but the 3rd and 6th
columns. At this stage the algorithm treats the possibilities di§erently: We
can place some ones before the puck (in the column indexes less than the puck
value), and/or some ones after the puck (in the column indexes greater than
the puck value). Before the puck, we can place a one where there is a 1, or
where there is a 2; after the puck, we can place a one in the 4th or 5th columns.
Suppose we place ones in the 4th and 5th columns. We drop the state vector
and advance the puck once more:
(2 1 2 0 1 0); 4
1
For the 4th round, we once again notice we can place some ones before the
puck, and/or some ones after.
Before the puck, we can place:
(a) two ones in columns of value 2 (1 choice)
(b) one one in the column of value 2 (2 choices)
(c) one one in the column of value 1 (1 choice)
(d) one one in a column of value 2 and one one in a column of value 1 (2
choices).
After we choose one of the options (a)-(d), we must multiply the listed
number of choices by one for each way to place any remaining ones to the right
of the puck.
So, for option (a), there is only one way to place the ones.
For option (b), there are two possible ways for each possible placement of
the remaining one to the right of the puck. Since there is only one nonzero value
remaining to the right of the puck, there are two ways total.
For option (c), there is one possible way for each possible placement of the
remaining one to the right of the puck. Again, since there is only one nonzero
value remaining, there is one way total.
For option (d), there are two possible ways to place the ones.
We choose option (a). We drop the state vector and advance the puck:
(1 1 1 0 1 0); 5
Since the puck is "covering" the 1 in the 5th column, we can only place
ones before the puck. There are (3 take 2) ways to place two ones in the three
columns of value 1, so we multiply 3 by the number of ways to get remaining
possibilities. After choosing the 1st and 3rd columns (though it doesnít matter
since weíre left of the puck; any two of the three will do), we drop the state
vector and advance the puck one final time:
(0 1 0 0 1 0); 6
There is only one way to place the ones in this situation, so we terminate
with a count of 1. But we must take into account all the multiplications along
the way: 1*1*1*1*3*1 = 3.
Another way of thinking of the varying row is to start with the first matrix,
focus on the lower-left 2x3 submatrix, and note how many ways there were to
permute the columns of that submatrix. Since there are only 3 such ways, we
get 3 matrices.
What I think I understand
This algorithm counts the the all possible 6x6 arrays with 2 1's in each row and column with none in the descending diagonal.
Instead of constructing the matrices it uses a "state_vector" filled with 6 2's, representing how many 2's are in that column, and a "puck" that represents the index of the diagonal and the current row as the algorithm iterates.
What I don't understand
The algorithm comes up with a value of 1 for each row except 5 which is assigned a 3, at the end these values are multiplied for the end result. These values are supposed to be the possible placements for each row but there are many possibilities for row 1, why was it given a one, why did the algorithm wait until row 5 to figure all the possible permutations?
Any help will be much appreciated!
I think what is going on is a tradeoff between doing combinatorics and doing recursion.
The algorithm is using recursion to add up all the counts for each choice of placing the 1's. The example considers a single choice at each stage, but to get the full count it needs to add the results for all possible choices.
Now it is quite possible to get the final answer simply using recursion all the way down. Every time we reach the bottom we just add 1 to the total count.
The normal next step is to cache the result of calling the recursive function as this greatly improves the speed. However, the memory use for such a dynamic programming approach depends on the number of states that need to be expanded.
The combinatorics in the later stages is making use of the fact that once the puck has passed a column, the exact arrangement of counts in the columns doesn't matter so you only need to evaluate one representative of each type and then add up the resulting counts multiplied by the number of equivalent ways.
This both reduces the memory use and improves the speed of the algorithm.
Note that you cannot use combinatorics for counts to the right of the puck, as for these the order of the counts is still important due to the restriction about the diagonal.
P.S. You can actually compute the number of ways for counting the number of n*n matrices with 2 1's in each column (and no diagonal entries) with pure combinatorics as:
a(n) = Sum_{k=0..n} Sum_{s=0..k} Sum_{j=0..n-k} (-1)^(k+j-s)*n!*(n-k)!*(2n-k-2j-s)!/(s!*(k-s)!*(n-k-j)!^2*j!*2^(2n-2k-j))
According to OEIS.
Given a 2-D array starting at (0,0) and proceeding to infinity in positive x and y axes. Given a number k>0 , find the number of cells reachable from (0,0) such that at every moment -> sum of digits of x+ sum of digits of y <=k . Moves can be up, down ,left or right. given x,y>=0 . Dfs gives answers but not sufficient for large values of k. anyone can help me with a better algorithm for this?
I think they asked you to calculate the number of cells (x,y) reachable with k>=x+y. If x=1 for example, then y can take any number between 0 and k-1 and the sum would be <=k. The total number of possibilities can be calculated by
sum(sum(1,y=0..k-x),x=0..k) = 1/2*k²+3/2*k+1
That should be able to do the trick for large k.
I am somewhat confused by the "digits" in your question. The digits make up the index like 3 times 9 makes 999. The sum of digits for the cell (999,888) would be 51. If you would allow the sum of digits to be 10^9 then you could potentially have 10^8 digits for an index, resulting something around 10^(10^8) entries, well beyond normal sizes for a table. I am therefore assuming my first interpretation. If that's not correct, then could you explain it a bit more?
EDIT:
okay, so my answer is not going to solve it. I'm afraid I don't see a nice formula or answer. I would approach it as a coloring/marking problem and mark all valid cells, then use some other technique to make sure all the parts are connected/to count them.
I have tried to come up with something but it's too messy. Basically I would try and mark large parts at once based on the index and k. If k=20, you can mark the cell range (0,0..299) at once (as any lower index will have a lower index sum) and continue to check the rest of the range. I start with 299 by fixing the 2 last digits to their maximum value and look for the max value for the first digit. Then continue that process for the remaining hundreds (300-999) and only fix the last digit to end up with 300..389 and 390..398. However, you can already see that it's a mess... (nevertheless i wanted to give it to you, you might get some better idea)
Another thing you can see immediately is that you problem is symmetric in index so any valid cell (x,y) tells you there's another valid cell (y,x). In a marking scheme / dfs/ bfs this can be exploited.
I need to randomly generate an NxN matrix of integers in the range 1 to K inclusive such that all rows and columns individually have the property that their elements are pairwise distinct.
For example for N=2 and K=3
This is ok:
1 2
2 1
This is not:
1 3
1 2
(Notice that if K < N this is impossible)
When K is sufficiently larger than N an efficient enough algorithm is just to generate a random matrix of 1..K integers, check that each row and each column is pairwise distinct, and if it isn't try again.
But what about the case where K is not much larger than N?
This is not a full answer, but a warning about an intuitive solution that does not work.
I am assuming that by "randomly generate" you mean with uniform probability on all existing such matrices.
For N=2 and K=3, here are the possible matrices, up to permutations of the set [1..K]:
1 2 1 2 1 2
2 1 2 3 3 1
(since we are ignoring permutations of the set [1..K], we can assume wlog that the first line is 1 2).
Now, an intuitive (but incorrect) strategy would be to draw the matrix entries one by one, ensuring for each entry that it is distinct from the other entries on the same line or column.
To see why it's incorrect, consider that we have drawn this:
1 2
x .
and we are now drawing x. x can be 2 or 3, but if we gave each possibility the probability 1/2, then the matrix
1 2
3 1
would get probability 1/2 of being drawn at the end, while it should have only probability 1/3.
Here is a (textual) solution. I don't think it provides good randomness, but nevertherless it could be ok for your application.
Let's generate a matrix in the range [0;K-1] (you will do +1 for all elements if you want to) with the following algorithm:
Generate the first line with any random method you want.
Each number will be the first element of a random sequence calculated in such a manner that you are guarranteed to have no duplicate in subsequent rows, that is for any distinct column x and y, you will have x[i]!=y[i] for all i in [0;N-1].
Compute each row for the previous one.
All the algorithm is based on the random generator with the property I mentioned. With a quick search, I found that the Inversive congruential generator meets this requirement. It seems to be easy to implement. It works if K is prime; if K is not prime, see on the same page 'Compound Inversive Generators'. Maybe it will be a little tricky to handle with perfect squares or cubic numbers (your problem sound like sudoku :-) ), but I think it is possible by creating compound generators with prime factors of K and different parametrization. For all generators, the first element of each column is the seed.
Whatever the value of K, the complexity is only depending on N and is O(N^2).
Deterministically generate a matrix having the desired property for rows and columns. Provided K > N, this can easily be done by starting the ith row with i, and filling in the rest of the row with i+1, i+2, etc., wrapping back to 1 after K. Other algorithms are possible.
Randomly permute columns, then randomly permute rows.
Let's show that permuting rows (i.e. picking up entire rows and assembling a new matrix from them in some order, with each row possibly in a different vertical position) leaves the desired properties intact for both rows and columns, assuming they were true before. The same reasoning then holds for column permutations, and for any sequence of permutations of either kind.
Trivially, permuting rows cannot change the property that, within each row, no element appears more than once.
The effect of permuting rows on a particular column is to reorder the elements within that column. This holds for any column, and since reordering elements cannot produce duplicate elements where there were none before, permuting rows cannot change the property that, within each column, no element appears more than once.
I'm not certain whether this algorithm is capable of generating all possible satisfying matrices, or if it does, whether it will generate all possible satisfying matrices with equal probability. Another interesting question that I don't have an answer for is: How many rounds of row-permutation-then-column-permutation are needed? More precisely, is any finite sequence of row-perm-then-column-perm rounds equivalent to a bounded number of (or in particular, one) row-perm-then-column-perm round? If so then nothing is gained by further permutations after the first row and column permutations. Perhaps someone with a stronger mathematics background can comment. But it may be good enough in any case.
Given an n×n matrix of real numbers. You are allowed to erase any number (from 0 to n) of rows and any number (from 0 to n) of columns, and after that the sum of the remaining entries is computed. Come up with an algorithm which finds out which rows and columns to erase in order to maximize that sum.
The problem is NP-hard. (So you should not expect a polynomial-time algorithm for solving this problem. There could still be (non-polynomial time) algorithms that are slightly better than brute-force, though.) The idea behind the proof of NP-hardness is that if we could solve this problem, then we could solve the the clique problem in a general graph. (The maximum-clique problem is to find the largest set of pairwise connected vertices in a graph.)
Specifically, given any graph with n vertices, let's form the matrix A with entries a[i][j] as follows:
a[i][j] = 1 for i == j (the diagonal entries)
a[i][j] = 0 if the edge (i,j) is present in the graph (and i≠j)
a[i][j] = -n-1 if the edge (i,j) is not present in the graph.
Now suppose we solve the problem of removing some rows and columns (or equivalently, keeping some rows and columns) so that the sum of the entries in the matrix is maximized. Then the answer gives the maximum clique in the graph:
Claim: In any optimal solution, there is no row i and column j kept for which the edge (i,j) is not present in the graph. Proof: Since a[i][j] = -n-1 and the sum of all the positive entries is at most n, picking (i,j) would lead to a negative sum. (Note that deleting all rows and columns would give a better sum, of 0.)
Claim: In (some) optimal solution, the set of rows and columns kept is the same. This is because starting with any optimal solution, we can simply remove all rows i for which column i has not been kept, and vice-versa. Note that since the only positive entries are the diagonal ones, we do not decrease the sum (and by the previous claim, we do not increase it either).
All of which means that if the graph has a maximum clique of size k, then our matrix problem has a solution with sum k, and vice-versa. Therefore, if we could solve our initial problem in polynomial time, then the clique problem would also be solved in polynomial time. This proves that the initial problem is NP-hard. (Actually, it is easy to see that the decision version of the initial problem — is there a way of removing some rows and columns so that the sum is at least k — is in NP, so the (decision version of the) initial problem is actually NP-complete.)
Well the brute force method goes something like this:
For n rows there are 2n subsets.
For n columns there are 2n subsets.
For an n x n matrix there are 22n subsets.
0 elements is a valid subset but obviously if you have 0 rows or 0 columns the total is 0 so there are really 22n-2+1 subsets but that's no different.
So you can work out each combination by brute force as an O(an) algorithm. Fast. :)
It would be quicker to work out what the maximum possible value is and you do that by adding up all the positive numbers in the grid. If those numbers happen to form a valid sub-matrix (meaning you can create that set by removing rows and/or columns) then there's your answer.
Implicit in this is that if none of the numbers are negative then the complete matrix is, by definition, the answer.
Also, knowing what the highest possible maximum is possibly allows you to shortcut the brute force evaluation since if you get any combination equal to that maximum then that is your answer and you can stop checking.
Also if all the numbers are non-positive, the answer is the maximum value as you can reduce the matrix to a 1 x 1 matrix with that 1 value in it, by definition.
Here's an idea: construct 2n-1 n x m matrices where 1 <= m <= n. Process them one after the other. For each n x m matrix you can calculate:
The highest possible maximum sum (as per above); and
Whether no numbers are positive allowing you to shortcut the answer.
if (1) is below the currently calculate highest maximum sum then you can discard this n x m matrix. If (2) is true then you just need a simple comparison to the current highest maximum sum.
This is generally referred to as a pruning technique.
What's more you can start by saying that the highest number in the n x n matrix is the starting highest maximum sum since obviously it can be a 1 x 1 matrix.
I'm sure you could tweak this into a (slightly more) efficient recursive tree-based search algorithm with the above tests effectively allowing you to eliminate (hopefully many) unnecessary searches.
We can improve on Cletus's generalized brute-force solution by modelling this as a directed graph. The initial matrix is the start node of the graph; its leaves are all the matrices missing one row or column, and so forth. It's a graph rather than a tree, because the node for the matrix without both the first column and row will have two parents - the nodes with just the first column or row missing.
We can optimize our solution by turning the graph into a tree: There's never any point exploring a submatrix with a column or row deleted that comes before the one we deleted to get to the current node, as that submatrix will be arrived at anyway.
This is still a brute-force search, of course - but we've eliminated the duplicate cases where we remove the same rows in different orders.
Here's an example implementation in Python:
def maximize_sum(m):
frontier = [(m, 0, False)]
best = None
best_score = 0
while frontier:
current, startidx, cols_done = frontier.pop()
score = matrix_sum(current)
if score > best_score or not best:
best = current
best_score = score
w, h = matrix_size(current)
if not cols_done:
for x in range(startidx, w):
frontier.append((delete_column(current, x), x, False))
startidx = 0
for y in range(startidx, h):
frontier.append((delete_row(current, y), y, True))
return best_score, best
And here's the output on 280Z28's example matrix:
>>> m = ((1, 1, 3), (1, -89, 101), (1, 102, -99))
>>> maximize_sum(m)
(106, [(1, 3), (1, 101)])
Since nobody asked for an efficient algorithm, use brute force: generate every possible matrix that can be created by removing rows and/or columns from the original matrix, choose the best one. A slightly more efficent version, which most likely can be proved to still be correct, is to generate only those variants where the removed rows and columns contain at least one negative value.
To try it in a simple way:
We need the valid subset of the set of entries {A00, A01, A02, ..., A0n, A10, ...,Ann} which max. sum.
First compute all subsets (the power set).
A valid subset is a member of the power set that for each two contained entries Aij and A(i+x)(j+y), contains also the elements A(i+x)j and Ai(j+y) (which are the remaining corners of the rectangle spanned by Aij and A(i+x)(j+y)).
Aij ...
. .
. .
... A(i+x)(j+y)
By that you can eliminate the invalid ones from the power set and find the one with the biggest sum in the remaining.
I'm sure it can be improved by improving an algorithm for power set generation in order to generate only valid subsets and by that avoiding step 2 (adjusting the power set).
I think there are some angles of attack that might improve upon brute force.
memoization, since there are many distinct sequences of edits that will arrive at the same submatrix.
dynamic programming. Because the search space of matrices is highly redundant, my intuition is that there would be a DP formulation that can save a lot of repeated work
I think there's a heuristic approach, but I can't quite nail it down:
if there's one negative number, you can either take the matrix as it is, remove the column of the negative number, or remove its row; I don't think any other "moves" result in a higher sum. For two negative numbers, your options are: remove neither, remove one, remove the other, or remove both (where the act of removal is either by axing the row or the column).
Now suppose the matrix has only one positive number and the rest are all <=0. You clearly want to remove everything but the positive entry. For a matrix with only 2 positive entries and the rest <= 0, the options are: do nothing, whittle down to one, whittle down to the other, or whittle down to both (resulting in a 1x2, 2x1, or 2x2 matrix).
In general this last option falls apart (imagine a matrix with 50 positives & 50 negatives), but depending on your data (few negatives or few positives) it could provide a shortcut.
Create an n-by-1 vector RowSums, and an n-by-1 vector ColumnSums. Initialize them to the row and column sums of the original matrix. O(n²)
If any row or column has a negative sum, remove edit: the one with the minimum such and update the sums in the other direction to reflect their new values. O(n)
Stop when no row or column has a sum less than zero.
This is an iterative variation improving on another answer. It operates in O(n²) time, but fails for some cases mentioned in other answers, which is the complexity limit for this problem (there are n² entries in the matrix, and to even find the minimum you have to examine each cell once).
Edit: The following matrix has no negative rows or columns, but is also not maximized, and my algorithm doesn't catch it.
1 1 3 goal 1 3
1 -89 101 ===> 1 101
1 102 -99
The following matrix does have negative rows and columns, but my algorithm selects the wrong ones for removal.
-5 1 -5 goal 1
1 1 1 ===> 1
-10 2 -10 2
mine
===> 1 1 1
Compute the sum of each row and column. This can be done in O(m) (where m = n^2)
While there are rows or columns that sum to negative remove the row or column that has the lowest sum that is less than zero. Then recompute the sum of each row/column.
The general idea is that as long as there is a row or a column that sums to nevative, removing it will result in a greater overall value. You need to remove them one at a time and recompute because in removing that one row/column you are affecting the sums of the other rows/columns and they may or may not have negative sums any more.
This will produce an optimally maximum result. Runtime is O(mn) or O(n^3)
I cannot really produce an algorithm on top of my head, but to me it 'smells' like dynamic programming, if it serves as a start point.
Big Edit: I honestly don't think there's a way to assess a matrix and determine it is maximized, unless it is completely positive.
Maybe it needs to branch, and fathom all elimination paths. You never no when a costly elimination will enable a number of better eliminations later. We can short circuit if it's found the theoretical maximum, but other than any algorithm would have to be able to step forward and back. I've adapted my original solution to achieve this behaviour with recursion.
Double Secret Edit: It would also make great strides to reduce to complexity if each iteration didn't need to find all negative elements. Considering that they don't change much between calls, it makes more sense to just pass their positions to the next iteration.
Takes a matrix, the list of current negative elements in the matrix, and the theoretical maximum of the initial matrix. Returns the matrix's maximum sum and the list of moves required to get there. In my mind move list contains a list of moves denoting the row/column removed from the result of the previous operation.
Ie: r1,r1
Would translate
-1 1 0 1 1 1
-4 1 -4 5 7 1
1 2 4 ===>
5 7 1
Return if sum of matrix is the theoretical maximum
Find the positions of all negative elements unless an empty set was passed in.
Compute sum of matrix and store it along side an empty move list.
For negative each element:
Calculate the sum of that element's row and column.
clone the matrix and eliminate which ever collection has the minimum sum (row/column) from that clone, note that action as a move list.
clone the list of negative elements and remove any that are effected by the action taken in the previous step.
Recursively call this algorithm providing the cloned matrix, the updated negative element list and the theoretical maximum. Append the moves list returned to the move list for the action that produced the matrix passed to the recursive call.
If the returned value of the recursive call is greater than the stored sum, replace it and store the returned move list.
Return the stored sum and move list.
I'm not sure if it's better or worse than the brute force method, but it handles all the test cases now. Even those where the maximum contains negative values.
This is an optimization problem and can be solved approximately by an iterative algorithm based on simulated annealing:
Notation: C is number of columns.
For J iterations:
Look at each column and compute the absolute benefit of toggling it (turn it off if it's currently on or turn it on if it's currently off). That gives you C values, e.g. -3, 1, 4. A greedy deterministic solution would just pick the last action (toggle the last column to get a benefit of 4) because it locally improves the objective. But that might lock us into a local optimum. Instead, we probabilistically pick one of the three actions, with probabilities proportional to the benefits. To do this, transform them into a probability distribution by putting them through a Sigmoid function and normalizing. (Or use exp() instead of sigmoid()?) So for -3, 1, 4 you get 0.05, 0.73, 0.98 from the sigmoid and 0.03, 0.42, 0.56 after normalizing. Now pick the action according to the probability distribution, e.g. toggle the last column with probability 0.56, toggle the second column with probability 0.42, or toggle the first column with the tiny probability 0.03.
Do the same procedure for the rows, resulting in toggling one of the rows.
Iterate for J iterations until convergence.
We may also, in early iterations, make each of these probability distributions more uniform, so that we don't get locked into bad decisions early on. So we'd raise the unnormalized probabilities to a power 1/T, where T is high in early iterations and is slowly decreased until it approaches 0. For example, 0.05, 0.73, 0.98 from above, raised to 1/10 results in 0.74, 0.97, 1.0, which after normalization is 0.27, 0.36, 0.37 (so it's much more uniform than the original 0.05, 0.73, 0.98).
It's clearly NP-Complete (as outlined above). Given this, if I had to propose the best algorithm I could for the problem:
Try some iterations of quadratic integer programming, formulating the problem as: SUM_ij a_ij x_i y_j, with the x_i and y_j variables constrained to be either 0 or 1. For some matrices I think this will find a solution quickly, for the hardest cases it would be no better than brute force (and not much would be).
In parallel (and using most of the CPU), use a approximate search algorithm to generate increasingly better solutions. Simulating Annealing was suggested in another answer, but having done research on similar combinatorial optimisation problems, my experience is that tabu search would find good solutions faster. This is probably close to optimal in terms of wandering between distinct "potentially better" solutions in the shortest time, if you use the trick of incrementally updating the costs of single changes (see my paper "Graph domination, tabu search and the football pool problem").
Use the best solution so far from the second above to steer the first by avoiding searching possibilities that have lower bounds worse than it.
Obviously this isn't guaranteed to find the maximal solution. But, it generally would when this is feasible, and it would provide a very good locally maximal solution otherwise. If someone had a practical situation requiring such optimisation, this is the solution that I'd think would work best.
Stopping at identifying that a problem is likely to be NP-Complete will not look good in a job interview! (Unless the job is in complexity theory, but even then I wouldn't.) You need to suggest good approaches - that is the point of a question like this. To see what you can come up with under pressure, because the real world often requires tackling such things.
yes, it's NP-complete problem.
It's hard to easily find the best sub-matrix,but we can easily to find some better sub-matrix.
Assume that we give m random points in the matrix as "feeds". then let them to automatically extend by the rules like :
if add one new row or column to the feed-matrix, ensure that the sum will be incrementive.
,then we can compare m sub-matrix to find the best one.
Let's say n = 10.
Brute force (all possible sets of rows x all possible sets of columns) takes
2^10 * 2^10 =~ 1,000,000 nodes.
My first approach was to consider this a tree search, and use
the sum of positive entries is an upper bound for every node in the subtree
as a pruning method. Combined with a greedy algorithm to cheaply generate good initial bounds, this yielded answers in about 80,000 nodes on average.
but there is a better way ! i later realised that
Fix some choice of rows X.
Working out the optimal columns for this set of rows is now trivial (keep a column if its sum of its entries in the rows X is positive, otherwise discard it).
So we can just brute force over all possible choices of rows; this takes 2^10 = 1024 nodes.
Adding the pruning method brought this down to 600 nodes on average.
Keeping 'column-sums' and incrementally updating them when traversing the tree of row-sets should allow the calculations (sum of matrix etc) at each node to be O(n) instead of O(n^2). Giving a total complexity of O(n * 2^n)
For slightly less than optimal solution, I think this is a PTIME, PSPACE complexity issue.
The GREEDY algorithm could run as follows:
Load the matrix into memory and compute row totals. After that run the main loop,
1) Delete the smallest row,
2) Subtract the newly omitted values from the old row totals
--> Break when there are no more negative rows.
Point two is a subtle detail: subtracted two rows/columns has time complexity n.
While re-summing all but two columns has n^2 time complexity!
Take each row and each column and compute the sum. For a 2x2 matrix this will be:
2 1
3 -10
Row(0) = 3
Row(1) = -7
Col(0) = 5
Col(1) = -9
Compose a new matrix
Cost to take row Cost to take column
3 5
-7 -9
Take out whatever you need to, then start again.
You just look for negative values on the new matrix. Those are values that actually substract from the overall matrix value. It terminates when there're no more negative "SUMS" values to take out (therefore all columns and rows SUM something to the final result)
In an nxn matrix that would be O(n^2)Log(n) I think
function pruneMatrix(matrix) {
max = -inf;
bestRowBitField = null;
bestColBitField = null;
for(rowBitField=0; rowBitField<2^matrix.height; rowBitField++) {
for (colBitField=0; colBitField<2^matrix.width; colBitField++) {
sum = calcSum(matrix, rowBitField, colBitField);
if (sum > max) {
max = sum;
bestRowBitField = rowBitField;
bestColBitField = colBitField;
}
}
}
return removeFieldsFromMatrix(bestRowBitField, bestColBitField);
}
function calcSumForCombination(matrix, rowBitField, colBitField) {
sum = 0;
for(i=0; i<matrix.height; i++) {
for(j=0; j<matrix.width; j++) {
if (rowBitField & 1<<i && colBitField & 1<<j) {
sum += matrix[i][j];
}
}
}
return sum;
}
Given a large sparse matrix (say 10k+ by 1M+) I need to find a subset, not necessarily continuous, of the rows and columns that form a dense matrix (all non-zero elements). I want this sub matrix to be as large as possible (not the largest sum, but the largest number of elements) within some aspect ratio constraints.
Are there any known exact or aproxamate solutions to this problem?
A quick scan on Google seems to give a lot of close-but-not-exactly results. What terms should I be looking for?
edit: Just to clarify; the sub matrix need not be continuous. In fact the row and column order is completely arbitrary so adjacency is completely irrelevant.
A thought based on Chad Okere's idea
Order the rows from largest count to smallest count (not necessary but might help perf)
Select two rows that have a "large" overlap
Add all other rows that won't reduce the overlap
Record that set
Add whatever row reduces the overlap by the least
Repeat at #3 until the result gets to small
Start over at #2 with a different starting pair
Continue until you decide the result is good enough
I assume you want something like this. You have a matrix like
1100101
1110101
0100101
You want columns 1,2,5,7 and rows 1 and 2, right? That submatrix would 4x2 with 8 elements. Or you could go with columns 1,5,7 with rows 1,2,3 which would be a 3x3 matrix.
If you want an 'approximate' method, you could start with a single non-zero element, then go on to find another non-zero element and add it to your list of rows and columns. At some point you'll run into a non-zero element that, if it's rows and columns were added to your collection, your collection would no longer be entirely non-zero.
So for the above matrix, if you added 1,1 and 2,2 you would have rows 1,2 and columns 1,2 in your collection. If you tried to add 3,7 it would cause a problem because 1,3 is zero. So you couldn't add it. You could add 2,5 and 2,7 though. Creating the 4x2 submatrix.
You would basically iterate until you can't find any more new rows and columns to add. That would get you too a local minimum. You could store the result and start again with another start point (perhaps one that didn't fit into your current solution).
Then just stop when you can't find any more after a while.
That, obviously, would take a long time, but I don't know if you'll be able to do it any more quickly.
I know you aren't working on this anymore, but I thought someone might have the same question as me in the future.
So, after realizing this is an NP-hard problem (by reduction to MAX-CLIQUE) I decided to come up with a heuristic that has worked well for me so far:
Given an N x M binary/boolean matrix, find a large dense submatrix:
Part I: Generate reasonable candidate submatrices
Consider each of the N rows to be a M-dimensional binary vector, v_i, where i=1 to N
Compute a distance matrix for the N vectors using the Hamming distance
Use the UPGMA (Unweighted Pair Group Method with Arithmetic Mean) algorithm to cluster vectors
Initially, each of the v_i vectors is a singleton cluster. Step 3 above (clustering) gives the order that the vectors should be combined into submatrices. So each internal node in the hierarchical clustering tree is a candidate submatrix.
Part II: Score and rank candidate submatrices
For each submatrix, calculate D, the number of elements in the dense subset of the vectors for the submatrix by eliminating any column with one or more zeros.
Select the submatrix that maximizes D
I also had some considerations regarding the min number of rows that needed to be preserved from the initial full matrix, and I would discard any candidate submatrices that did not meet this criteria before selecting a submatrix with max D value.
Is this a Netflix problem?
MATLAB or some other sparse matrix libraries might have ways to handle it.
Is your intent to write your own?
Maybe the 1D approach for each row would help you. The algorithm might look like this:
Loop over each row
Find the index of the first non-zero element
Find the index of the non-zero row element with the largest span between non-zero columns in each row and store both.
Sort the rows from largest to smallest span between non-zero columns.
At this point I start getting fuzzy (sorry, not an algorithm designer). I'd try looping over each row, lining up the indexes of the starting point, looking for the maximum non-zero run of column indexes that I could.
You don't specify whether or not the dense matrix has to be square. I'll assume not.
I don't know how efficient this is or what its Big-O behavior would be. But it's a brute force method to start with.
EDIT. This is NOT the same as the problem below.. My bad...
But based on the last comment below, it might be equivilent to the following:
Find the furthest vertically separated pair of zero points that have no zero point between them.
Find the furthest horizontally separated pair of zero points that have no zeros between them ?
Then the horizontal region you're looking for is the rectangle that fits between these two pairs of points?
This exact problem is discussed in a gem of a book called "Programming Pearls" by Jon Bentley, and, as I recall, although there is a solution in one dimension, there is no easy answer for the 2-d or higher dimensional variants ...
The 1=D problem is, effectively, find the largest sum of a contiguous subset of a set of numbers:
iterate through the elements, keeping track of a running total from a specific previous element, and the maximum subtotal seen so far (and the start and end elemnt that generateds it)... At each element, if the maxrunning subtotal is greater than the max total seen so far, the max seen so far and endelemnt are reset... If the max running total goes below zero, the start element is reset to the current element and the running total is reset to zero ...
The 2-D problem came from an attempt to generate a visual image processing algorithm, which was attempting to find, within a stream of brightnesss values representing pixels in a 2-color image, find the "brightest" rectangular area within the image. i.e., find the contained 2-D sub-matrix with the highest sum of brightness values, where "Brightness" was measured by the difference between the pixel's brighness value and the overall average brightness of the entire image (so many elements had negative values)
EDIT: To look up the 1-D solution I dredged up my copy of the 2nd edition of this book, and in it, Jon Bentley says "The 2-D version remains unsolved as this edition goes to print..." which was in 1999.