getrf_batch (Group Version)¶
Computes the batch of LU factorizations of a batch of general m-by-n matrices.
This routine belongs to the oneapi::mkl::lapack
namespace.
Description¶
The routine computes the LU factorizations of a batch of general
m
-by-n
matrices A
i, (iϵ{1...batch_size}
) as
Where P
i is a permutation matrix, L
i is
lower triangular with unit diagonal elements (lower trapezoidal if
m > n
) and U
i is upper triangular (upper trapezoidal
if m < n
). The routine uses partial pivoting with row
interchanges.
Total number of problems to solve, batch_size
, is a sum of sizes
of all of the groups of parameters as provided by
group_sizes
array.
API¶
Syntax¶
namespace oneapi::mkl::lapack {
sycl::event getrf_batch(sycl::queue &queue,
std::int64_t *m,
std::int64_t *n,
T **a,
std::int64_t *lda,
std::int64_t **ipiv,
std::int64_t group_count,
std::int64_t *group_sizes,
T *scratchpad,
std::int64_t scratchpad_size,
const std::vector<sycl::event> &events = {})
}
This function supports the following precisions and devices:
T |
Devices supported |
---|---|
|
CPU and GPU |
|
CPU and GPU |
|
CPU and GPU |
|
CPU and GPU |
Input Parameters¶
- queue
Device queue where calculations will be performed.
- m
Array of
group_count
parameters mg specifying the number of rows in the matricesA
i belonging to groupg
(m
g ≥ 0).- n
Array of
group_count
parameters ng specifying the number of columns in the matricesA
i belonging to groupg
(n
g ≥ 0).- a
Array of
batch_size
pointers to input matricesA
i. The pointers must be device-accessible.- lda
Array of
group_count
parameters ldag specifying the leading dimension ofA
i belonging to groupg
(lda
g ≥max(1, m
g)).- group_count
Specifies the number of groups of parameters. Must be at least 0.
- group_sizes
Array of group_count integers. Array element with index
g
specifies the number of problems to solve for each of the groups of parametersg
. So the total number of problems to solve,batch_size
, is a sum of all parameter group sizes.- scratchpad
Scratchpad memory to be used by routine for storing intermediate results.
- scratchpad_size
Size of scratchpad memory as a number of floating point elements of type T. Size should not be less then the value returned by getrf_batch_scratchpad_size (Group Version).
- events
List of events to wait for before starting computation. Defaults to empty list.
Output Parameters¶
- a
Matrices pointed to by array
a
are overwritten byL
i andU
i. The unit diagonal elements ofL
i are not stored.- ipiv
Arrays of batch_size pointers to arrays containing pivot indices
ipiv
i each of size at leastmax(1,min(m
g,n
g)); for 1 ≤k
≤min(m
g,n
g), rowk
ofA
i was interchanged with rowipiv
i(k
). The pointers must be device-accessible.
Exceptions¶
Exception |
Description |
---|---|
|
This exception is thrown when problems occur during calculations. You can obtain the info code of the problem using the info() method of the exception object: If If If |
Return Values¶
Output event to wait on to ensure computation is complete.