I have an N by 2 matrix A of indices of elements I want to get from a 2D matrix B, each row of A being the row and column index of an element of B that I want to get. I would like to get all of those elements stacked up as an N by 1 vector.
B is a square matrix, so I am currently using
N = size(B,1);
indices = arrayfun(#(i) A(i,1) + N*(A(i,2)-1), 1:size(A,1));
result = B(indices);
but, while it works, this is probing to be a huge bottleneck and I need to speed up the code in order for it to be useful.
What is the fastest way I can achieve the same result?
How about
indices = [1 N] * (A'-1) + 1;
I can never remember if B(A(:,1), A(:,2)) works the way you want it to, but I'd try that to avoid the intermediate variable. If that does not work, try subs2ind.
Also, you can look at how you generated A in the first place. if A came about from the output of find, for example, it is faster to use logical indexing. i.e if
B( B == 2 )
Is faster than finding the row,col indexes that satisfy that condition, then indexing into B.
Related
I want to know if their are any fast approaches to solve the following problem. I have a list of codes somewhere in the thousands (A0, A1, A2, ...). There is a positive value attached to about a million distinct combinations (A0-A1, A2-A10, A1-A2-A10, ...). Let the values be denoted f(A0-A1). Note that not all the combinations have the value attached.
For each listed combination, I want to calculate the sum of values of the values attached to each set that contains the given combination. For instance, for A2-A10,
calculate
g(A2-A10) = f(A2-A10) + f(A1-A2-A10) + ...
I would like to do this with minimal time complexity. A simpler related problem is to find all combinations where g(C) is greater than a threshold value.
Key the existing combinations with a bit map, where bit n denotes whether An is in that particular coding. Store the values keyed by the bit map for each in your favorite hash-map structure. Thus, f(A0, A1, A10, A12) would be combo_val[11000000001010000...]
To sum all of the desired combinations, build a bit map of your root. For instance, with the combination above, we'd have root = 1100000000101000 (cutting off at 16 total elements for the sake of illustration.
Now simply loop through the keys of the hashmap, using root as a mask. Sum the desired values:
total = 0
for key in combo_val.keys()
if root && key == root
total += combo_val[key]
Does that get you moving?
I thought waaay too long before coming up with the following approach.
Index the million combinations. So you know which you want. In your example:
0: A0-A1
1: A2-A10
2: A1-A2-A10
For each code, create an ordered list of combinations that contain that code. Call that code_combs. In your example:
A0: [0]
A1: [0, 2]
A2: [1, 2]
A10: [1, 2]
Now we have a combination of codes, like A2-A10. We create two arrays, one of codes, the other of indices. Set indices at 0. So:
codes = ['A2', 'A10']
indices = [0, 0]
And now do the following:
while not done:
let max_comb = max(code_combs[codes[i]][indices[i]] over i in range(len(codes))
Advance each index until we are at the max_comb or greater
(if we reach the end of any list, we are done)
If all are at the same max_comb, we add its value.
Advance all indexes by 1.
(if we reach the end of any list, we are done)
Basically this is a k-way intersection of ordered lists. Now here is the trick. If we advance naively, this will be slightly faster because we only have to look at combinations that contain a code. However we can use a clever advance strategy like this:
Advance by 1, 2, 4, 8, etc until we reach or pass the point we want.
Do a binary search between the last two values until we find the point we want
(Be warned, implementing binary search is not always so easy to get right.)
And now we are crossing fingers. But if any one of our codes has few combinations that it is in, and there aren't too many codes in our combination, we can compute our intersection quite quickly.
I'm working with arrays of integer, all of the same size l.
I have a static set of them and I need to build a function to efficiently look them up.
The tricky part is that the elements in the array I need to search might be off by 1.
Given the arrays {A_1, A_2, ..., A_n}, and an array S, I need a function search such that:
search(S)=x iff ∀i: A_x[i] ∈ {S[i]-1, S[i], S[i]+1}.
A possible solution is treating each vector as a point in an l-dimensional space and looking for the closest point, but it'd cost something like O(l*n) in space and O(l*log(n)) in time.
Would there be a solution with a better space complexity (and/or time, of course)?
My arrays are pretty different from each other, and good heuristics might be enough.
Consider a search array S with the values:
S = [s1, s2, s3, ... , sl]
and the average value:
s̅ = (s1 + s2 + s3 + ... + sl) / l
and two matching arrays, one where every value is one greater than the corresponding value in S, and one where very value is one smaller:
A1 = [s1+1, s2+1, s3+1, ... , sl+1]
A2 = [s1−1, s2−1, s3−1, ... , sl−1]
These two arrays would have the average values:
a̅1 = (s1 + 1 + s2 + 1 + s3 + 1 + ... + sl + 1) / l = s̅ + 1
a̅2 = (s1 − 1 + s2 − 1 + s3 − 1 + ... + sl − 1) / l = s̅ − 1
So every matching array, whose values are at most 1 away from the corresponding values in the search array, has an average value that is at most 1 away from the average value of the search array.
If you calculate and store the average value of each array, and then sort the arrays based on their average value (or use an extra data structure that enables you to find all arrays with a certain average value), you can quickly identify which arrays have an average value within 1 of the search array's average value. Depending on the data, this could drastically reduce the number of arrays you have to check for similarity.
After having pre-processed the arrays and stores their average values, performing a search would mean iterating over the search array to calculate the average value, looking up which arrays have a similar average value, and then iterating over those arrays to check every value.
If you expect many arrays to have a similar average value, you could use several averages to detect arrays that are locally very different but similar on average. You could e.g. calculate these four averages:
the first half of the array
the second half of the array
the odd-numbered elements
the even-numbered elements
Analysis of the actual data should give you more information about how to divide the array and combine different averages to be most effective.
If the total sum of an array cannot exceed the integer size, you could store the total sum of each array, and check whether it is within l of the total sum of the search array, instead of using averages. This would avoid having to use floats and divisions.
(You could expand this idea by also storing other properties which are easily calculated and don't take up much space to store, such as the highest and lowest value, the biggest jump, ... They could help create a fingerprint of each array that is near-unique, depending on the data.)
If the number of dimensions is not very small, then probably the best solution will be to build a decision tree that recursively partitions the set along different dimensions.
Each node, including the root, would be a hash table from the possible values for some dimension to either:
The list of points that match that value within tolerance, if it's small enough; or
Those same points in a similar tree partitioning on the remaining dimensions.
Since each level completely eliminates one dimension, the depth of the tree is at most L, and search takes O(L) time.
The order in which the dimensions are chosen along each path is important, of course -- the wrong choice could explode the size of the data structure, with each point appearing many times.
Since your points are "pretty different", though, it should be possible to build a tree with minimal duplication. I would try the ID3 algorithm to choose the dimensions: https://en.wikipedia.org/wiki/ID3_algorithm. That basically means you greedily choose the dimension that maximizes the overall reduction in set size, using an entropy metric.
I would personally create something like a Trie for the lookup. I said "something like" because we have up to 3 values per index that might match. So we aren't creating a decision tree, but a DAG. Where sometimes we have choices.
That is straightforward and will run (with backtracking) in maximum time O(k*l).
But here is the trick. Whenever we see a choice of matching states that we can go into next, we can create a merged state which tries all of them. We can create a few or a lot of these merged states. Each one will defer a choice by 1 step. And if we're careful to keep track of which merged states we've created, we can reuse the same one over and over again.
In theory we can be generating partial matches for somewhat arbitrary subsets of our arrays. Which can grow exponentially in the number of arrays. In practice are likely to only wind up with a few of these merged states. But still we can guarantee a tradeoff - more states up front runs faster later. So we optimize until we are done or have hit the limit of how much data we want to have.
Here is some proof of concept code for this in Python. It will likely build the matcher in time O(n*l) and match in time O(l). However it is only guaranteed to build the matcher in time O(n^2 * l^2) and match in time O(n * l).
import pprint
class Matcher:
def __init__ (self, arrays, optimize_limit=None):
# These are the partial states we could be in during a match.
self.states = [{}]
# By state, this is what we would be trying to match.
self.state_for = ['start']
# By combination we could try to match for, which state it is.
self.comb_state = {'start': 0}
for i in range(len(arrays)):
arr = arrays[i]
# Set up "matched the end".
state_index = len(self.states)
this_state = {'matched': [i]}
self.comb_state[(i, len(arr))] = state_index
self.states.append(this_state)
self.state_for.append((i, len(arr)))
for j in reversed(range(len(arr))):
this_for = (i, j)
prev_state = {}
if 0 == j:
prev_state = self.states[0]
matching_values = set((arr[k] for k in range(max(j-1, 0), min(j+2, len(arr)))))
for v in matching_values:
if v in prev_state:
prev_state[v].append(state_index)
else:
prev_state[v] = [state_index]
if 0 < j:
state_index = len(self.states)
self.states.append(prev_state)
self.state_for.append(this_for)
self.comb_state[this_for] = state_index
# Theoretically optimization can take space
# O(2**len(arrays) * len(arrays[0]))
# We will optimize until we are done or hit a more reasonable limit.
if optimize_limit is None:
# Normally
optimize_limit = len(self.states)**2
# First we find all of the choices at the root.
# This will be an array of arrays with format:
# [state, key, values]
todo = []
for k, v in self.states[0].iteritems():
if 1 < len(v):
todo.append([self.states[0], k, tuple(v)])
while len(todo) and len(self.states) < optimize_limit:
this_state, this_key, this_match = todo.pop(0)
if this_key == 'matched':
pass # We do not need to optimize this!
elif this_match in self.comb_state:
this_state[this_key] = self.comb_state[this_match]
else:
# Construct a new state that is all of these.
new_state = {}
for state_ind in this_match:
for k, v in self.states[state_ind].iteritems():
if k in new_state:
new_state[k] = new_state[k] + v
else:
new_state[k] = v
i = len(self.states)
self.states.append(new_state)
self.comb_state[this_match] = i
self.state_for.append(this_match)
this_state[this_key] = [i]
for k, v in new_state.iteritems():
if 1 < len(v):
todo.append([new_state, k, tuple(v)])
#pp = pprint.PrettyPrinter()
#pp.pprint(self.states)
#pp.pprint(self.comb_state)
#pp.pprint(self.state_for)
def match (self, list1, ind=0, state=0):
this_state = self.states[state]
if 'matched' in this_state:
return this_state['matched']
elif list1[ind] in this_state:
answer = []
for next_state in this_state[list1[ind]]:
answer = answer + self.match(list1, ind+1, next_state)
return answer;
else:
return []
foo = Matcher([[1, 2, 3], [2, 3, 4]])
print(foo.match([2, 2, 3]))
Please note that I deliberately set up a situation where there are 2 matches. It reports both of them. :-)
I came up with a further approach derived off Matt Timmermans's answer: building a simple decision tree that might have certain some arrays in multiple branches. It works even if the error in the array I'm searching is larger than 1.
The idea is the following: given the set of arrays As...
Pick an index and a pivot.
I fixed the pivot to a constant value that works well with my data, and tried all indices to find the best one. Trying multiple pivots might work better, but I didn't need to.
Partition As into two possibly-intersecting subsets, one for the arrays (whose index-th element is) smaller than the pivot, one for the larger arrays. Arrays very close to the pivot are added to both sets:
function partition( As, pivot, index ):
return {
As.filter( A => A[index] <= pivot + 1 ),
As.filter( A => A[index] >= pivot - 1 ),
}
Apply both previous steps to each subset recursively, stopping when a subset only contains a single element.
Here an example of a possible tree generated with this algorithm (note that A2 appears both on the left and right child of the root node):
{A1, A2, A3, A4}
pivot:15
index:73
/ \
/ \
{A1, A2} {A2, A3, A4}
pivot:7 pivot:33
index:54 index:0
/ \ / \
/ \ / \
A1 A2 {A2, A3} A4
pivot:5
index:48
/ \
/ \
A2 A3
The search function then uses this as a normal decision tree: it starts from the root node and recurses either to the left or the right child depending on whether its value at index currentNode.index is greater or less than currentNode.pivot. It proceeds recursively until it reaches a leaf.
Once the decision tree is built, the time complexity is in the worst case O(n), but in practice it's probably closer to O(log(n)) if we choose good indices and pivots (and if the dataset is diverse enough) and find a fairly balanced tree.
The space complexity can be really bad in the worst case (O(2^n)), but it's closer to O(n) with balanced trees.
I am working on something which I feel an NP-hard problem. So, I am not looking for the optimal solution but I am looking for a better heuristics. An integer input matrix (matrix A in the following example) is given as input and I have to produce an integer output matrix (matrix B in the following example) whose number of rows are smaller than the input matrix and should obey the following two conditions:
1) Each column of the output matrix should contain integers in the same order as they appear in the input matrix. (In the example below, first column of the matrix A and matrix B have the same integers 1,3 in the same order.)
2) Same integers must not appear in the same row (In the example below, first row of the matrix B contains the integers 1,3 and 2 which are different from each other.)
Note that the input matrix always obey the 2nd condition.
What a greedy algorithm looks like to solve this problem?
Example:
In this example the output matrix 'Matrix B' contains all the integers as they appear in the input matrix 'Matrix A" but the output matrix has 5 rows and the input matrix has 6 rows. So, the output 'Matrix B' is a valid solution of the input 'Matrix A'.
I would produce the output one row at a time. When working out what to put in the row I would consider the next number from each input column, starting from the input column which has the most numbers yet to be placed, and considering the columns in decreasing order of numbers yet to be placed. Where a column can put a number in the current output row when its turn comes up it should do so.
You could extend this to a branch and bound solution to find the exact best answer. Recursively try all possible rows at each stage, except when you can see that the current row cannot possibly improve on the best answer so far. You know that if you have a column with k entries yet to be placed, in the best possible case you will need at least k more rows.
In practice I would expect that this will be too expensive to be practical, so you will need to ignore some possible rows which you cannot rule out, and so cannot guarantee to find the best answer. You could try using a heuristic search such as Limited Discrepancy search.
Another non-exact speedup is to multiply the estimate for the number of rows that the best possible answer derived from a partial solution will require by some factor F > 1. This will allow you to rule out some solutions earlier than branch and bound. The answer you find can be no more than F times more expensive than the best possible answer, because you only discard possibilities that cannot improve on the current answer by more than a factor of F.
A greedy solution to this problem would involve placing the numbers column by column, top down, as they appear.
Pseudocode:
For each column c in A:
r = 0 // row index of next element in A
nextRow = 0 // row index of next element to be placed in B
while r < A.NumRows()
while r < A.NumRows() && A[r, c] is null:
r++ // increment row to check in A
if r < A.NumRows() // we found a non-null entry in A
while nextRow < A.NumRows() && ~CheckConstraints(A[r,c], B[nextRow, c]):
nextRow++ // increment output row in B
if 'nextRow' >= A.NumRows()
return unsolvable // couldn't find valid position in B
B[nextRow, c] = v // successfully found position in B
++nextRow // increment output row in B
If there are no conflicts you end up "packing" B as tightly as possible. Otherwise you greedily search for the next non-conflicting row position in B. If none can be found, the problem is unsolvable.
The helper function CheckConstraints checks backwards in columns for the same row value in B to ensure the same value hasn't already been placed in a row.
If the problem statement is relaxed such that the output row count in B is <= the row count in A, then if we are unable to pack B any tighter, then we can return A as a solution.
I have a sparse matrix and I need to fill certain entries with a specific value, I am using a for loop right now but I know its not the correct way to do it so I was wondering if its possible to vectorise this for loop?
K = sparse(N);
for i=vectorofrandomintegers
K(i,i) = 1;
end
If I vectorise it normally as so:
K(A,A) = 1;
then it fills all the entries in each row denoted by A whereas I want individual entries (i.e. K(1,1) = 1 or K(6,6)=1).
Also, the entries are not diagonally adjacent so I can't plop the identity matrix into it.
If you are going to use a vectorized method, you would need to get the linear indices to be set. The issue is that if you define your sparse matrix as K = sparse(N) and then linearly index into K, it would extend the size of it in one direction only and not along both row and column. Thus, you need to specify to MATLAB that you are
looking to use this sparse to store a 2D array. Thus, it would be -
K = sparse(N,N);
Get the linear indices to index into K using sub2ind and set them -
ind1 = sub2ind([N N],vectorofrandomintegers,vectorofrandomintegers);
K(ind1) = 1;
It's fairly simple
i'd use
K((A-1)*N+A))=1;
i believe that should fix your problem by treating the matrix as a vector
Instead of declaring and then filling a sparse matrix, you can fill it at the same time you define it:
i = vectorofrandomintegers; j = i;
K = sparse(i,j,1,N,N)
I am a new student learning to use Matlab.
Could anyone please tell me is there a faster way possibly without loops:
to assign for each row only two values 1, -1 into different positions of a big sparse matrix.
My code to build a bimatrix or bibimatrix for the MILP problem of condition :
f^k_{ij} <= y_{ij} for every arc (i,j) and all k ~=r; in a multi-commodity flow model.
Naive approach:
bimatrix=[];
% create each row and then add to bimatrix
newrow4= zeros(1,n*(n+1)^2);
for k=1:n
for i=0:n
for j=1: n
if j~=i
%change value of some positions to -1 and 1
newrow4(i*n^2+(j-1)*n+k)=1;
newrow4((n+1)*n^2+i*n+j)=-1;
% add to bimatrix
bimatrix=[bimatrix; newrow4];
% change newrow4 back to zeros row.
newrow4(i*n^2+(j-1)*n+k)=0;
newrow4((n+1)*n^2+i*n+j)=0;
end
end
end
end
OR:
% Generate the big sparse matrix first.
bibimatrix=zeros(n^3 ,n*(n+1)^2);
t=1;
for k=1:n
for i=0:n
for j=1: n
if j~=i
%Change 2 positions in each row to -1 and 1 in each row.
bibimatrix(t,i*n^2+(j-1)*n+k)=1;
bibimatrix(t,(n+1)*n^2+i*n+j)=-1;
t=t+1
end
end
end
end
With these above code in Matlab, the time to generate this matrix, with n~12, is more than 3s. I need to generate a larger matrix in less time.
Thank you.
Suggestion: Use sparse matrices.
You should be able to create two vectors containing the column number where you want your +1 and -1 in each row. Let's call these two vectors vec_1 and vec_2. You should be able to do this without loops (if not, I still think the procedure below will be faster).
Let the size of your matrix be (max_row X max_col). Then you can create your matrix like this:
bibimatrix = sparse(1:max_row,vec_1,1,max_row,max_col);
bibimatrix = bibimatrix + sparse(1:max_row, vec_2,-1,max_row,max_col)
If you want to see the entire matrix (which you don't, since it's huge) you can write: full(bibimatrix).
EDIT:
You may also do it this way:
col_vec = [vec_1, vec_2];
row_vec = [1:max_row, 1:max_row];
s = [ones(1,max_row), -1*ones(1,max_row)];
bibimatrix = sparse(row_vec, col_vec, s, max_row, max_col)
Disclaimer: I don't have MATLAB available, so it might not be error-free.