# 0.4 Implementing kronecker products efficiently  (Page 2/2)

 Page 2 / 2
$\mathcal{C}\left(g\right)={\mathcal{Q}}_{g\left(1\right)}{c}_{g\left(2\right)}{c}_{g\left(3\right)}+{r}_{g\left(1\right)}{\mathcal{Q}}_{g\left(2\right)}{c}_{g\left(3\right)}+{r}_{g\left(1\right)}{r}_{g\left(2\right)}{\mathcal{Q}}_{g\left(3\right)}$

In general

$\mathcal{C}\left(g\right)=\sum _{i=1}^{n}\left(\prod _{j=1}^{i-1},{r}_{g\left(j\right)}\right){\mathcal{Q}}_{g\left(i\right)}\left(\prod _{j=i+1}^{n},{c}_{g\left(j\right)}\right).$

Therefore, the most efficient factorization of ${\otimes }_{i}{A}_{i}$ is described by the permutation $g\left(·\right)$ that minimizes $\mathcal{C}$ .

It turns out that for the Kronecker product of more than two matrices, the ordering of operations that describesthe most efficient factorization of ${\otimes }_{i}{A}_{i}$ also depends only on the ratios $\left({r}_{i}-{c}_{i}\right)/{\mathcal{Q}}_{i}$ . To show that this is the case, suppose $u\left(·\right)$ is the permutation that minimizes $\mathcal{C}$ , then $u\left(·\right)$ has the property that

$\frac{{r}_{u\left(k\right)}-{c}_{u\left(k\right)}}{{\mathcal{Q}}_{u\left(k\right)}}\le \frac{{r}_{u\left(k+1\right)}-{c}_{u\left(k+1\right)}}{{\mathcal{Q}}_{u\left(k+1\right)}}$

for $k=1,\cdots ,n-1$ . To support this, note that since $u\left(·\right)$ is the permutation that minimizes $\mathcal{C}$ , we have in particular

$\mathcal{C}\left(u\right)\le \mathcal{C}\left(v\right)$

where $v\left(·\right)$ is the permutation defined by the following:

$v\left(i\right)=\left\{\begin{array}{cc}u\left(i\right)\hfill & ik+1\hfill \\ u\left(k+1\right)\hfill & i=k\hfill \\ u\left(k\right)\hfill & i=k+1\hfill \end{array}.$

Because only two terms in [link] are different, we have from [link]

$\sum _{i=k}^{k+1}\left(\prod _{j=1}^{i-1},{r}_{u\left(j\right)}\right){\mathcal{Q}}_{u\left(i\right)}\left(\prod _{j=i+1}^{n},{c}_{u\left(j\right)}\right)\le \sum _{i=k}^{k+1}\left(\prod _{j=1}^{i-1},{r}_{v\left(j\right)}\right){\mathcal{Q}}_{v\left(i\right)}\left(\prod _{j=i+1}^{n},{c}_{v\left(j\right)}\right)$

which, after canceling common terms from each side, gives

${\mathcal{Q}}_{u\left(k\right)}{c}_{u\left(k+1\right)}+{r}_{u\left(k\right)}{\mathcal{Q}}_{u\left(k+1\right)}\le {\mathcal{Q}}_{v\left(k\right)}{c}_{v\left(k+1\right)}+{r}_{v\left(k\right)}{\mathcal{Q}}_{v\left(k+1\right)}.$

Since $v\left(k\right)=u\left(k+1\right)$ and $v\left(k+1\right)=u\left(k\right)$ this becomes

${\mathcal{Q}}_{u\left(k\right)}{c}_{u\left(k+1\right)}+{r}_{u\left(k\right)}{\mathcal{Q}}_{u\left(k+1\right)}\le {\mathcal{Q}}_{u\left(k+1\right)}{c}_{u\left(k\right)}+{r}_{u\left(k+1\right)}{\mathcal{Q}}_{u\left(k\right)}$

which is equivalent to [link] . Therefore, to find the best factorization of ${\otimes }_{i}{A}_{i}$ it is necessary only to compute the ratios $\left({r}_{i}-{c}_{i}\right)/{\mathcal{Q}}_{i}$ and to order them in an non-decreasing order. The operation ${A}_{i}$ whose index appears first in this list is applied to the data vector $x$ first, and so on

As above, if ${r}_{u\left(k+1\right)}>{c}_{u\left(k+1\right)}$ and ${r}_{u\left(k\right)}<{c}_{u\left(k\right)}$ then [link] is always true. Therefore, in the most computationally efficientfactorization of ${\otimes }_{i}{A}_{i}$ , all matrices with fewer rows than columns are always applied to the data vector $x$ before any matrices with more rows than columns.If some matrices are square, then their ordering does not affect the computational efficiency as longas they are applied after all matrices with fewer rows than columns and before all matrices with more rows than columns.

Once the permutation $g\left(·\right)$ that minimizes $\mathcal{C}$ is determined by ordering the ratios $\left({r}_{i}-{c}_{i}\right)/{\mathcal{Q}}_{i}$ , ${\otimes }_{i}{A}_{i}$ can be written as

$\underset{i=1}{\overset{n}{\otimes }}{A}_{i}=\prod _{i=n}^{1}{I}_{a\left(i\right)}\otimes {A}_{g\left(i\right)}\otimes {I}_{b\left(i\right)}$

where

$a\left(i\right)=\prod _{k=1}^{g\left(i\right)-1}\gamma \left(i,k\right)$
$b\left(i\right)=\prod _{k=g\left(i\right)+1}^{n}\gamma \left(i,k\right)$

and where $\gamma \left(·\right)$ is defined by

$\gamma \left(i,k\right)=\left\{\begin{array}{cc}{r}_{k}\hfill & \text{if}\phantom{\rule{4.pt}{0ex}}g\left(i\right)>g\left(k\right)\hfill \\ {c}_{k}\hfill & \text{if}\phantom{\rule{4.pt}{0ex}}g\left(i\right)

## Some matlab code

A Matlab program that computes the permutation that describes the computationally most efficientfactorization of ${\otimes }_{i=1}^{n}{A}_{i}$ is cgc() . It also gives the resulting computational cost.It requires the computational cost of each of the matrices ${A}_{i}$ and the number of rows and columns of each.

function [g,C] = cgc(Q,r,c,n)% [g,C] = cgc(Q,r,c,n);% Compute g and C % g : permutation that minimizes C% C : computational cost of Kronecker product of A(1),...,A(n) % Q : computation cost of A(i)% r : rows of A(i) % c : columns of A(i)% n : number of terms f = find(Q==0);Q(f) = eps * ones(size(Q(f))); Q = Q(:);r = r(:); c = c(:);[s,g] = sort((r-c)./Q);C = 0; for i = 1:nC = C + prod(r(g(1:i-1)))*Q(g(i))*prod(c(g(i+1:n))); endC = round(C);

The Matlab program kpi() implements the Kronecker product ${\otimes }_{i=1}^{n}{A}_{i}$ .

function y = kpi(d,g,r,c,n,x) % y = kpi(d,g,r,c,n,x);% Kronecker Product : A(d(1)) kron ... kron A(d(n)) % g : permutation of 1,...,n% r : [r(1),...,r(n)] % c : [c(1),..,c(n)]% r(i) : rows of A(d(i)) % c(i) : columns of A(d(i))% n : number of terms for i = 1:na = 1; for k = 1:(g(i)-1)if i>find(g==k) a = a * r(k);else a = a * c(k);end endb = 1; for k = (g(i)+1):nif i>find(g==k) b = b * r(k);else b = b * c(k);end end% y = (I(a) kron A(d(g(i))) kron I(b)) * x; y = IAI(d(g(i)),a,b,x);end

where the last line of code calls a function that implements $\left({I}_{a}\otimes {A}_{d\left(g\left(i\right)\right)}\otimes {I}_{b}\right)x$ . That is, the program IAI(i,a,b,x) implements $\left({I}_{a}\otimes A\left(i\right)\otimes {I}_{b}\right)x$ .

The Matlab program IAI implements $y=\left({I}_{m}\otimes A\otimes {I}_{n}\right)x$

function y = IAI(A,r,c,m,n,x) % y = (I(m) kron A kron I(n))x% r : number of rows of A % c : number of columns of Av = 0:n:n*(r-1); u = 0:n:n*(c-1);for i = 0:m-1 for j = 0:n-1y(v+i*r*n+j+1) = A * x(u+i*c*n+j+1); endend

It simply uses two loops to implement the $mn$ copies of $A$ . Each copy of $A$ is applied to a different subset of the elements of $x$ .

## Vector/parallel interpretation

The command $I\otimes A\otimes I$ where $\otimes$ is the Kronecker (or Tensor) product can be interpreted as avector/parallel command [link] , [link] . In these references, the implementation of thesecommands is discussed in detail and they have found that the Tensor product is“an extremely useful tool for matching algorithms to computer architectures [link] .”

The expression $I\otimes A$ can easily be seen to represent a parallel command:

$I\otimes A=\left[\begin{array}{cccc}A& & & \\ & A& & \\ & & \ddots & \\ & & & A\end{array}\right].$

Each block along the diagonal acts on non-overlapping sections of the data vector - so that each sectioncan be performed in parallel. Since each section represents exactly the same operation, this form isamenable to implementation on a computer with aparallel architectural configuration. The expression $A\otimes I$ can be similarly seen to represent a vector command, see [link] .

It should also be noted that by employing `stride' permutations, the command $\left(I\otimes A\otimes I\right)x$ can be replaced by either $\left(I\otimes A\right)x$ or $\left(A\otimes I\right)x$ [link] , [link] . It is only necessary to permute the input and output.It is also the case that these stride permutations are natural loading and storing commands for some architectures.

In the programs we have written in conjunction with this paper we implement the commands $y=\left(I\otimes A\otimes I\right)x$ with loops in a set of subroutines. The circular convolution and prime length FFT programswe present, however, explicitly use the form $I\otimes A\otimes I$ to make clear the structure of the algorithm, to make themmore modular and simpler, and to make them amenable to implementation on special architectures.In fact, in [link] it is suggested that it might be practical to develop tensor product compilers.The FFT programs we have generated will be well suited for such compilers.

Application of nanotechnology in medicine
what is variations in raman spectra for nanomaterials
I only see partial conversation and what's the question here!
what about nanotechnology for water purification
please someone correct me if I'm wrong but I think one can use nanoparticles, specially silver nanoparticles for water treatment.
Damian
yes that's correct
Professor
I think
Professor
what is the stm
is there industrial application of fullrenes. What is the method to prepare fullrene on large scale.?
Rafiq
industrial application...? mmm I think on the medical side as drug carrier, but you should go deeper on your research, I may be wrong
Damian
How we are making nano material?
what is a peer
What is meant by 'nano scale'?
What is STMs full form?
LITNING
scanning tunneling microscope
Sahil
how nano science is used for hydrophobicity
Santosh
Do u think that Graphene and Fullrene fiber can be used to make Air Plane body structure the lightest and strongest. Rafiq
Rafiq
what is differents between GO and RGO?
Mahi
what is simplest way to understand the applications of nano robots used to detect the cancer affected cell of human body.? How this robot is carried to required site of body cell.? what will be the carrier material and how can be detected that correct delivery of drug is done Rafiq
Rafiq
if virus is killing to make ARTIFICIAL DNA OF GRAPHENE FOR KILLED THE VIRUS .THIS IS OUR ASSUMPTION
Anam
analytical skills graphene is prepared to kill any type viruses .
Anam
what is Nano technology ?
write examples of Nano molecule?
Bob
The nanotechnology is as new science, to scale nanometric
brayan
nanotechnology is the study, desing, synthesis, manipulation and application of materials and functional systems through control of matter at nanoscale
Damian
Is there any normative that regulates the use of silver nanoparticles?
what king of growth are you checking .?
Renato
What fields keep nano created devices from performing or assimulating ? Magnetic fields ? Are do they assimilate ?
why we need to study biomolecules, molecular biology in nanotechnology?
?
Kyle
yes I'm doing my masters in nanotechnology, we are being studying all these domains as well..
why?
what school?
Kyle
biomolecules are e building blocks of every organics and inorganic materials.
Joe
anyone know any internet site where one can find nanotechnology papers?
research.net
kanaga
sciencedirect big data base
Ernesto
Introduction about quantum dots in nanotechnology
hi
Loga
what does nano mean?
nano basically means 10^(-9). nanometer is a unit to measure length.
Bharti
how did you get the value of 2000N.What calculations are needed to arrive at it
Privacy Information Security Software Version 1.1a
Good
Got questions? Join the online conversation and get instant answers!