org.ejml.alg.dense.misc

Class TransposeAlgs

java.lang.Object

org.ejml.alg.dense.misc.TransposeAlgs

public class TransposeAlgs
extends java.lang.Object

Low level transpose algorithms. No sanity checks are performed. Take a look at BenchmarkTranspose to see which one is faster on your computer.

Constructor Summary

Constructors

Constructor and Description
TransposeAlgs()

Method Summary

Modifier and Type	Method and Description
static void	block(org.ejml.data.RowD1Matrix64F A, org.ejml.data.RowD1Matrix64F A_tran, int blockLength) Performs a transpose across block sub-matrices.
static void	square(org.ejml.data.RowD1Matrix64F mat) In-place transpose for a square matrix.
static void	standard(org.ejml.data.RowD1Matrix64F A, org.ejml.data.RowD1Matrix64F A_tran) A straight forward transpose.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

TransposeAlgs

public TransposeAlgs()

Method Detail

square

public static void square(org.ejml.data.RowD1Matrix64F mat)

In-place transpose for a square matrix. On most architectures it is faster than the standard transpose algorithm, but on most modern computers it's slower than block transpose.

Parameters:

mat - The matrix that is transposed in-place. Modified.

block

public static void block(org.ejml.data.RowD1Matrix64F A,
                         org.ejml.data.RowD1Matrix64F A_tran,
                         int blockLength)

Performs a transpose across block sub-matrices. Reduces the number of cache misses on larger matrices. *NOTE* If this is beneficial is highly dependent on the computer it is run on. e.g: - Q6600 Almost twice as fast as standard. - Pentium-M Same speed and some times a bit slower than standard.

Parameters:

A - Original matrix. Not modified.

A_tran - Transposed matrix. Modified.

blockLength - Length of a block.

standard

public static void standard(org.ejml.data.RowD1Matrix64F A,
                            org.ejml.data.RowD1Matrix64F A_tran)

A straight forward transpose. Good for small non-square matrices.

Parameters:

A - Original matrix. Not modified.

A_tran - Transposed matrix. Modified.