pqc/external/flint-2.4.3/fmpq_mat/doc/fmpq_mat.txt
2014-05-24 23:16:06 +02:00

429 lines
16 KiB
Plaintext

/*=============================================================================
This file is part of FLINT.
FLINT is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
FLINT is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with FLINT; if not, write to the Free Software
Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
=============================================================================*/
/******************************************************************************
Copyright (C) 2011 Fredrik Johansson
******************************************************************************/
*******************************************************************************
Memory management
*******************************************************************************
void fmpq_mat_init(fmpq_mat_t mat, slong rows, slong cols)
Initialises a matrix with the given number of rows and columns for use.
void fmpq_mat_clear(fmpq_mat_t mat)
Frees all memory associated with the matrix. The matrix must be
reinitialised if it is to be used again.
*******************************************************************************
Entry access
*******************************************************************************
MACRO fmpq_mat_entry(mat,i,j)
Gives a reference to the entry at row \code{i} and column \code{j}.
The reference can be passed as an input or output variable to any
\code{fmpq} function for direct manipulation of the matrix element.
No bounds checking is performed.
MACRO fmpq_mat_entry_num(mat,i,j)
Gives a reference to the numerator of the entry at row \code{i} and
column \code{j}. The reference can be passed as an input or output
variable to any \code{fmpz} function for direct manipulation of the
matrix element. No bounds checking is performed.
MACRO fmpq_mat_entry_den(mat,i,j)
Gives a reference to the denominator of the entry at row \code{i} and
column \code{j}. The reference can be passed as an input or output
variable to any \code{fmpz} function for direct manipulation of the
matrix element. No bounds checking is performed.
*******************************************************************************
Basic assignment
*******************************************************************************
void fmpq_mat_set(fmpq_mat_t dest, const fmpq_mat_t src)
Sets the entries in \code{dest} to the same values as in \code{src},
assuming the two matrices have the same dimensions.
void fmpq_mat_zero(fmpq_mat_t mat)
Sets \code{mat} to the zero matrix.
void fmpq_mat_one(fmpq_mat_t mat)
Let $m$ be the minimum of the number of rows and columns
in the matrix \code{mat}. This function sets the first
$m \times m$ block to the identity matrix, and the remaining
block to zero.
void fmpq_mat_transpose(fmpq_mat_t rop, const fmpq_mat_t op)
Sets the matrix \code{rop} to the tranpose of the matrix \code{op},
assuming that their dimensios are compatible.
*******************************************************************************
Addition, scalar multiplication
*******************************************************************************
void fmpq_mat_add(fmpq_mat_t mat,
const fmpq_mat_t mat1, const fmpq_mat_t mat2)
Sets \code{mat} to the sum of \code{mat1} and \code{mat2},
assuming that all three matrices have the same dimensions.
void fmpq_mat_sub(fmpq_mat_t mat,
const fmpq_mat_t mat1, const fmpq_mat_t mat2)
Sets \code{mat} to the difference of \code{mat1} and \code{mat2},
assuming that all three matrices have the same dimensions.
void fmpq_mat_neg(fmpq_mat_t rop, const fmpq_mat_t op)
Sets \code{rop} to the negative of \code{op}, assuming that
the two matrices have the same dimensions.
void fmpq_mat_scalar_mul_fmpz(fmpq_mat_t rop,
const fmpq_mat_t op, const fmpz_t x)
Sets \code{rop} to \code{op} multiplied by the integer $x$,
assuming that the two matrices have the same dimensions.
Note that the integer $x$ may not be aliased with any part of
the entries of \code{rop}.
void fmpq_mat_scalar_div_fmpz(fmpq_mat_t rop,
const fmpq_mat_t op, const fmpz_t x)
Sets \code{rop} to \code{op} divided by the integer $x$,
assuming that the two matrices have the same dimensions
and that $x$ is non-zero.
Note that the integer $x$ may not be aliased with any part of
the entries of \code{rop}.
*******************************************************************************
Input and output
*******************************************************************************
void fmpq_mat_print(const fmpq_mat_t mat)
Prints the matrix \code{mat} to standard output.
*******************************************************************************
Random matrix generation
*******************************************************************************
void fmpq_mat_randbits(fmpq_mat_t mat, flint_rand_t state, mp_bitcnt_t bits)
This is equivalent to applying \code{fmpq_randbits} to all entries
in the matrix.
void fmpq_mat_randtest(fmpq_mat_t mat, flint_rand_t state, mp_bitcnt_t bits)
This is equivalent to applying \code{fmpq_randtest} to all entries
in the matrix.
*******************************************************************************
Special matrices
*******************************************************************************
void fmpq_mat_hilbert_matrix(fmpq_mat_t mat)
Sets \code{mat} to a Hilbert matrix of the given size. That is,
the entry at row $i$ and column $j$ is set to $1/(i+j+1)$.
*******************************************************************************
Basic comparison and properties
*******************************************************************************
int fmpq_mat_equal(const fmpq_mat_t mat1, const fmpq_mat_t mat2)
Returns nonzero if \code{mat1} and \code{mat2} have the same shape and
all their entries agree, and returns zero otherwise. Assumes the
entries in both \code{mat1} and \code{mat2} are in canonical form.
int fmpq_mat_is_integral(const fmpq_mat_t mat)
Returns nonzero if all entries in \code{mat} are integer-valued, and
returns zero otherwise. Assumes that the entries in \code{mat}
are in canonical form.
int fmpq_mat_is_zero(const fmpq_mat_t mat)
Returns nonzero if all entries in \code{mat} are zero, and returns
zero otherwise.
int fmpq_mat_is_empty(const fmpq_mat_t mat)
Returns a non-zero value if the number of rows or the number of
columns in \code{mat} is zero, and otherwise returns
zero.
int fmpq_mat_is_square(const fmpq_mat_t mat)
Returns a non-zero value if the number of rows is equal to the
number of columns in \code{mat}, and otherwise returns zero.
*******************************************************************************
Integer matrix conversion
*******************************************************************************
int fmpq_mat_get_fmpz_mat(fmpz_mat_t dest, const fmpq_mat_t mat)
Sets \code{dest} to \code{mat} and returns nonzero if all entries
in \code{mat} are integer-valued. If not all entries in \code{mat}
are integer-valued, sets \code{dest} to an undefined matrix
and returns zero. Assumes that the entries in \code{mat} are
in canonical form.
void fmpq_mat_get_fmpz_mat_entrywise(fmpz_mat_t num, fmpz_mat_t den,
const fmpq_mat_t mat)
Sets the integer matrices \code{num} and \code{den} respectively
to the numerators and denominators of the entries in \code{mat}.
void fmpq_mat_get_fmpz_mat_matwise(fmpz_mat_t num, fmpz_t den,
const fmpq_mat_t mat)
Converts all entries in \code{mat} to a common denominator,
storing the rescaled numerators in \code{num} and the
denominator in \code{den}. The denominator will be minimal
if the entries in \code{mat} are in canonical form.
void fmpq_mat_get_fmpz_mat_rowwise(fmpz_mat_t num, fmpz * den,
const fmpq_mat_t mat)
Clears denominators in \code{mat} row by row. The rescaled
numerators are written to \code{num}, and the denominator
of row \code{i} is written to position \code{i} in \code{den}
which can be a preinitialised \code{fmpz} vector. Alternatively,
\code{NULL} can be passed as the \code{den} variable, in which
case the denominators will not be stored.
void fmpq_mat_get_fmpz_mat_rowwise_2(fmpz_mat_t num, fmpz_mat_t num2,
fmpz * den, const fmpq_mat_t mat, const fmpq_mat_t mat2)
Clears denominators row by row of both \code{mat} and \code{mat2},
writing the respective numerators to \code{num} and \code{num2}.
This is equivalent to concatenating \code{mat} and \code{mat2}
horizontally, calling \code{fmpq_mat_get_fmpz_mat_rowwise},
and extracting the two submatrices in the result.
void fmpq_mat_get_fmpz_mat_colwise(fmpz_mat_t num, fmpz * den,
const fmpq_mat_t mat)
Clears denominators in \code{mat} column by column. The rescaled
numerators are written to \code{num}, and the denominator
of column \code{i} is written to position \code{i} in \code{den}
which can be a preinitialised \code{fmpz} vector. Alternatively,
\code{NULL} can be passed as the \code{den} variable, in which
case the denominators will not be stored.
void fmpq_mat_set_fmpz_mat(fmpq_mat_t dest, const fmpz_mat_t src)
Sets \code{dest} to \code{src}.
void fmpq_mat_set_fmpz_mat_div_fmpz(fmpq_mat_t mat, const fmpz_mat_t num,
const fmpz_t den)
Sets \code{mat} to the integer matrix \code{num} divided by the
common denominator \code{den}.
*******************************************************************************
Modular reduction and rational reconstruction
*******************************************************************************
void fmpq_mat_get_fmpz_mat_mod_fmpz(fmpz_mat_t dest, const fmpq_mat_t mat,
const fmpz_t mod)
Sets each entry in \code{dest} to the corresponding entry in \code{mat},
reduced modulo \code{mod}.
int fmpq_mat_set_fmpz_mat_mod_fmpz(fmpq_mat_t X, const fmpz_mat_t Xmod,
const fmpz_t mod)
Set \code{X} to the entrywise rational reconstruction integer matrix
\code{Xmod} modulo \code{mod}, and returns nonzero if the reconstruction
is successful. If rational reconstruction fails for any element,
returns zero and sets the entries in \code{X} to undefined values.
*******************************************************************************
Matrix multiplication
*******************************************************************************
void fmpq_mat_mul_direct(fmpq_mat_t C, const fmpq_mat_t A, const fmpq_mat_t B)
Sets \code{C} to the matrix product \code{AB}, computed
naively using rational arithmetic. This is typically very slow and
should only be used in circumstances where clearing denominators
would consume too much memory.
void fmpq_mat_mul_cleared(fmpq_mat_t C, const fmpq_mat_t A,
const fmpq_mat_t B)
Sets \code{C} to the matrix product \code{AB}, computed
by clearing denominators and multiplying over the integers.
void fmpq_mat_mul(fmpq_mat_t C, const fmpq_mat_t A, const fmpq_mat_t B)
Sets \code{C} to the matrix product \code{AB}. This
simply calls \code{fmpq_mat_mul_cleared}.
void fmpq_mat_mul_fmpz_mat(fmpq_mat_t C, const fmpq_mat_t A,
const fmpz_mat_t B)
Sets \code{C} to the matrix product \code{AB}, with \code{B}
an integer matrix. This function works efficiently by clearing
denominators of \code{A}.
void fmpq_mat_mul_r_fmpz_mat(fmpq_mat_t C, const fmpz_mat_t A,
const fmpq_mat_t B)
Sets \code{C} to the matrix product \code{AB}, with \code{A}
an integer matrix. This function works efficiently by clearing
denominators of \code{B}.
*******************************************************************************
Trace
*******************************************************************************
void fmpq_mat_trace(fmpq_t trace, const fmpq_mat_t mat)
Computes the trace of the matrix, i.e. the sum of the entries on
the main diagonal. The matrix is required to be square.
*******************************************************************************
Determinant
*******************************************************************************
void fmpq_mat_det(fmpq_t det, const fmpq_mat_t mat)
Sets \code{det} to the determinant of \code{mat}. In the general case,
the determinant is computed by clearing denominators and computing a
determinant over the integers. Matrices of size 0, 1 or 2 are handled
directly.
*******************************************************************************
Nonsingular solving
*******************************************************************************
int fmpq_mat_solve_fraction_free(fmpq_mat_t X, const fmpq_mat_t A,
const fmpq_mat_t B)
Solves \code{AX = B} for nonsingular \code{A} by clearing denominators
and solving the rescaled system over the integers using a fraction-free
algorithm. This is usually the fastest algorithm for small systems.
Returns nonzero if \code{X} is nonsingular or if the right hand side
is empty, and zero otherwise.
int fmpq_mat_solve_dixon(fmpq_mat_t X, const fmpq_mat_t A, const fmpq_mat_t B)
Solves \code{AX = B} for nonsingular \code{A} by clearing denominators
and solving the rescaled system over the integers using Dixon's algorithm.
The rational solution matrix is generated using rational reconstruction.
This is usually the fastest algorithm for large systems.
Returns nonzero if \code{X} is nonsingular or if the right hand side
is empty, and zero otherwise.
*******************************************************************************
Inverse
*******************************************************************************
int fmpq_mat_inv(fmpq_mat_t B, const fmpq_mat_t A)
Sets \code{B} to the inverse matrix of \code{A} and returns nonzero.
Returns zero if \code{A} is singular. \code{A} must be a square matrix.
*******************************************************************************
Echelon form
*******************************************************************************
int fmpq_mat_pivot(slong * perm, fmpq_mat_t mat, slong r, slong c)
Helper function for row reduction. Returns 1 if the entry of \code{mat}
at row $r$ and column $c$ is nonzero. Otherwise searches for a nonzero
entry in the same column among rows $r+1, r+2, \ldots$. If a nonzero
entry is found at row $s$, swaps rows $r$ and $s$ and the corresponding
entries in \code{perm} (unless \code{NULL}) and returns -1. If no
nonzero pivot entry is found, leaves the inputs unchanged and returns 0.
slong fmpq_mat_rref_classical(fmpq_mat_t B, const fmpq_mat_t A)
Sets \code{B} to the reduced row echelon form of \code{A} and returns
the rank. Performs Gauss-Jordan elimination directly over the rational
numbers. This algorithm is usually inefficient and is mainly intended
to be used for testing purposes.
slong fmpq_mat_rref_fraction_free(fmpq_mat_t B, const fmpq_mat_t A)
Sets \code{B} to the reduced row echelon form of \code{A} and returns
the rank. Clears denominators and performs fraction-free Gauss-Jordan
elimination using \code{fmpz_mat} functions.
slong fmpq_mat_rref(fmpq_mat_t B, const fmpq_mat_t A)
Sets \code{B} to the reduced row echelon form of \code{A} and returns
the rank. This function automatically chooses between the classical and
fraction-free algorithms depending on the size of the matrix.