NAG CL Interface
f10dac (randproj_dct_real)
1
Purpose
f10dac computes a fast random projection of a real $m$ by $n$ matrix $A$ using a Discrete Cosine Transform (DCT). The function can be used as a building block in Randomised Numerical Linear Algebra (RNLA) algorithms, such as $QR$ decompositions, Singular Value Decompositions (SVDs), and (approximate) least squares solvers.
2
Specification
void 
f10dac (Nag_TransType trans,
Integer m,
Integer n,
double a[],
Integer pda,
Integer k,
Integer state[],
NagError *fail) 

The function may be called by the names: f10dac or nag_rnla_randproj_dct_real.
3
Description
A random projection is written as either
or
where
$A$ is a real
$m$ by
$n$ matrix,
$\Omega $ is an
$n$ by
$k$ matrix in the first case, and a
$k$ by
$m$ matrix in the second case. These cases are referred to as random projection by postmultiplication and random projection by premultiplication, respectively.
When a random projection by postmultiplication uses the DCT, it is written as
where
$D$ is a diagonal matrix whose values are uniformly distributed on the set
$\left\{1,\mathrm{+1}\right\}$,
$F$ is a DCT, and
$C$ is a matrix that selects a subset of columns, uniformly at random.
When a random projection by premultiplication uses the DCT, it is written as
The operators
$D$ and
$F$ are as above and
$R$ is a matrix that selects a subset of rows, again uniformly at random.
None of these matrix operators require a full matrixmatrix product to be computed. The computational complexity of applying this type of random projection is
$\mathit{O}\left(mn\mathrm{log}\left(k\right)\right)$. More details of the DCTbased random projection can be found in
Avron et al. (2010).
The DCTbased random projection is closely related to the Subsampled Random Fourier Transform (SRFT) presented in Section 4.6 of
Halko et al. (2011).
4
References
Avron H, Maymounkov P and Toledo S (2010) Blendenpik: Supercharging LAPACK's leastsquares solver SIAM J. Sci. Comput. 32(3) 1217–1236
Halko N (2012) Randomized methods for computing lowrank approximations of matrices PhD thesis
Halko N, Martinsson P G and Tropp J A (2011) Finding structure with randomness: Probabilistic algorithms for constructing approximate matrix decompositions
SIAM Rev. 53(2) 217–288
https://epubs.siam.org/doi/abs/10.1137/090771806
5
Arguments

1:
$\mathbf{trans}$ – Nag_TransType
Input

On entry: specifies whether the operation premultiplies
$A$ or postmultiplies
$A$.
 ${\mathbf{trans}}=\mathrm{Nag\_NoTrans}$
 Random projection is done by postmultiplication, $Y=A\Omega $.
 ${\mathbf{trans}}=\mathrm{Nag\_Trans}$
 Random projection is done by premultiplication, $Y=\Omega A$.
Constraint:
${\mathbf{trans}}=\mathrm{Nag\_NoTrans}$ or $\mathrm{Nag\_Trans}$.

2:
$\mathbf{m}$ – Integer
Input

On entry: $m$, the number of rows of the matrix $A$.
Constraint:
${\mathbf{m}}>0$.

3:
$\mathbf{n}$ – Integer
Input

On entry: $n$, the number of columns of the matrix $A$.
Constraint:
${\mathbf{n}}>0$.

4:
$\mathbf{a}\left[\mathit{dim}\right]$ – double
Input/Output

Note: the dimension,
dim, of the array
a
must be at least
${\mathbf{pda}}\times {\mathbf{n}}$.
The $\left(i,j\right)$th element of the matrix $A$ is stored in ${\mathbf{a}}\left[\left(j1\right)\times {\mathbf{pda}}+i1\right]$.
On entry: the $m$ by $n$ matrix $A$.
On exit: if
${\mathbf{trans}}=\mathrm{Nag\_NoTrans}$, the first
$k$ columns of
$A$ are overwritten with the
$m$ by
$k$ random projection of
$A$.
If ${\mathbf{trans}}=\mathrm{Nag\_Trans}$, the first $k$ rows of $A$ are overwritten with the $k$ by $n$ random projection of $A$.

5:
$\mathbf{pda}$ – Integer
Input

On entry: the stride separating matrix row elements in the array
a.
Constraint:
${\mathbf{pda}}\ge {\mathbf{m}}$.

6:
$\mathbf{k}$ – Integer
Input

On entry: $k$, number of columns in the random projection, $Y=A\Omega $.
Constraints:
 if ${\mathbf{trans}}=\mathrm{Nag\_NoTrans}$, $0<{\mathbf{k}}\le {\mathbf{n}}$;
 if ${\mathbf{trans}}=\mathrm{Nag\_Trans}$, $0<{\mathbf{k}}\le {\mathbf{m}}$.

7:
$\mathbf{state}\left[\mathit{dim}\right]$ – Integer
Communication Array

Note: the dimension,
$\mathit{dim}$, of this array is dictated by the requirements of associated functions that must have been previously called. This array MUST be the same array passed as argument
state in the previous call to
nag_rand_init_repeatable (g05kfc) or
nag_rand_init_nonrepeatable (g05kgc).
On entry: contains information on the selected base generator and its current state.
On exit: contains updated information on the state of the generator.

8:
$\mathbf{fail}$ – NagError *
Input/Output

The NAG error argument (see
Section 7 in the Introduction to the NAG Library CL Interface).
6
Error Indicators and Warnings
 NE_ALLOC_FAIL

Dynamic memory allocation failed.
See
Section 3.1.2 in the Introduction to the NAG Library CL Interface for further information.
 NE_BAD_PARAM

On entry, argument $\u2329\mathit{\text{value}}\u232a$ had an illegal value.
 NE_INT

On entry, ${\mathbf{k}}=\u2329\mathit{\text{value}}\u232a$ and ${\mathbf{trans}}=\u2329\mathit{\text{value}}\u232a$.
Constraint: if ${\mathbf{trans}}=\mathrm{Nag\_NoTrans}$, $0<{\mathbf{k}}\le {\mathbf{n}}$, if ${\mathbf{trans}}=\mathrm{Nag\_Trans}$, $0<{\mathbf{k}}\le {\mathbf{m}}$.
On entry, ${\mathbf{m}}=\u2329\mathit{\text{value}}\u232a$.
Constraint: ${\mathbf{m}}>0$.
On entry, ${\mathbf{n}}=\u2329\mathit{\text{value}}\u232a$.
Constraint: ${\mathbf{n}}>0$.
On entry, ${\mathbf{pda}}=\u2329\mathit{\text{value}}\u232a$.
Constraint: ${\mathbf{pda}}\ge {\mathbf{m}}$.
On entry,
state vector has been corrupted or not initialized.
 NE_INTERNAL_ERROR

An internal error has occurred in this function. Check the function call and any array sizes. If the call is correct then please contact
NAG for assistance.
See
Section 7.5 in the Introduction to the NAG Library CL Interface for further information.
 NE_NO_LICENCE

Your licence key may have expired or may not have been installed correctly.
See
Section 8 in the Introduction to the NAG Library CL Interface for further information.
7
Accuracy
The accuracy of algorithms that use
f10dac depend on the extent to which the random projection,
$Y$, captures the range (i.e., column space) of the matrix
$A$. More formally, the following probabilistic error bound holds,
with failure probability at most
$\mathit{O}\left({k}^{1}\right)$, and where
${\sigma}_{k+1}$ is the
$\left(k+1\right)$th singular value and the norm on the lefthand side of the equation is the spectral norm.
The matrix $\left(I{P}_{Y}\right)$ is sometimes referred to as the orthogonal projector onto the complementary subspace, ${\mathrm{range}\left({P}_{Y}\right)}^{\perp}$.
Informally,
$Y$ captures the range of
$A$ well if
is dominated by the first term,
${P}_{Y}A$.
See Section 8–11 of
Halko et al. (2011) for more details.
8
Parallelism and Performance
f10dac is threaded by NAG for parallel execution in multithreaded implementations of the NAG Library.
f10dac makes calls to BLAS and/or LAPACK routines, which may be threaded within the vendor library used by this implementation. Consult the documentation for the vendor library for further information.
Please consult the
X06 Chapter Introduction for information on how to control and interrogate the OpenMP environment used within this function. Please also consult the
Users' Note for your implementation for any additional implementationspecific information.
f10dac uses the same DCT as
c06rfc. The time taken by
c06rfc is fastest if the only prime factors of the projected dimension are
$2$,
$3$ or
$5$. In order to make the performance of
f10dac less sensitive to the size of the projected dimension, the input,
a, is padded with zeros up to the nearest power of
$2$ or the nearest
$1000$, whichever is smaller.
10
Example
This example applies a random projection using the DCT to the
$6$ by
$6$ matrix
It then computes a
$6$ by
$5$ orthornormal matrix
$Q$ and a probabilistic error bound,
$E$ that quantifies how well the range of
$Q$ approximates the range of
$A$. The matrix
$Q$ is calculated through a
$QR$ factorization of the projection
$Y=A\Omega $. The error bound is based on the following result from Section 4.3 of
Halko et al. (2011). Given a sequence
$\left\{{\omega}^{\left(i\right)}:i=1,2,\dots ,r\right\}$ of standard Gaussian vectors, where
$r$ is some positive integer, then the measure,
$E$, of the error in how well the projection captures the range of
$A$ is given by
where
$\Vert \xb7\Vert $ is the spectral norm when applied to a matrix and the Euclidean norm when applied to a vector.
The exact approximation error $\Vert \left(IQ{Q}^{*}\right)A\Vert $ is also computed for comparison.
This example is constructed so that the true rank of $A$ is $4$: the last $3$ rows are identical so the matrix only has $4$ linearly independent rows. This means that when $k\ge 4$ the accuracy of $Q$ will be close to machine precision.
10.1
Program Text
10.2
Program Data
10.3
Program Results