blob: 8e507bed40216064c99d4cd16d6fa9898bad64db [file] [log] [blame]
2007-02-17 Brian Gough <bjg@network-theory.co.uk>
* test_source.c (FUNCTION): avoid running negative value tests on
char, because it can be unsigned.
2007-01-26 Brian Gough <bjg@network-theory.co.uk>
* minmax_source.c: added support for NaNs
2006-10-31 Brian Gough <bjg@network-theory.co.uk>
* prop_source.c (FUNCTION): added functions gsl_matrix_ispos,
gsl_matrix_isneg
2003-01-01 Brian Gough <brian.gough@network-theory.co.uk>
* gsl_matrix_complex_float.h (gsl_matrix_complex_float_get):
removed const from zero
* matrix_source.c (FUNCTION): removed const from zero
2002-11-24 Brian Gough <bjg@network-theory.co.uk>
* Makefile.am: added libgslsys.a to link, to provide gsl_hypot for
complex division
Mon Jun 17 22:31:33 2002 Brian Gough <bjg@network-theory.co.uk>
* test_complex_source.c (FUNCTION): fixed non-constant initializer
Wed May 1 21:33:44 2002 Brian Gough <bjg@network-theory.co.uk>
* gsl_matrix_complex_float.h (gsl_matrix_complex_float_get): moved
constant zero inside GSL_RANGE_CHECK_OFF
Sun Mar 24 20:28:48 2002 Brian Gough <bjg@network-theory.co.uk>
* oper_complex_source.c: complex matrix operations (from Toby
White)
Mon Feb 18 20:33:58 2002 Brian Gough <bjg@network-theory.co.uk>
* copy_source.c (gsl_matrix_swap): fixed prototype by removing
const from second arg.
Sun Jan 27 22:29:37 2002 Brian Gough <bjg@network-theory.co.uk>
* test.c: ensure that range check is working when running the
tests
Fri Sep 14 18:56:34 2001 Brian Gough <bjg@network-theory.co.uk>
* view_source.c: fixed cast for array type
* view.c: added #includes for missing const qualified types
* view_source.c: error for non-unit strides
Fri Aug 3 14:11:23 2001 Brian Gough <bjg@network-theory.co.uk>
* added gsl_matrix_ptr and gsl_matrix_const_ptr functions
Mon Jul 16 21:28:05 2001 Brian Gough <bjg@network-theory.co.uk>
* rowcol_source.c (FUNCTION): initialized view to NULL
Fri Jul 13 21:29:27 2001 Brian Gough <bjg@network-theory.co.uk>
* changed views to be structs and used casts to initialize them
Mon Jul 2 12:35:16 2001 Brian Gough <bjg@network-theory.co.uk>
* view.h: provide macros for initializing null
vectors and views
Sun Jul 1 22:44:51 2001 Brian Gough <bjg@network-theory.co.uk>
* introduction of new-style vector/matrix views
Fri Jun 1 17:04:52 2001 Brian Gough <bjg@network-theory.co.uk>
* getset_source.c: made these routines work with the current
matrix struct, previously they would give incorrect results
Mon May 14 22:43:39 2001 Brian Gough <bjg@network-theory.co.uk>
* matrix_source.c (FUNCTION): removed unnecessary inline from
static function definition
Tue May 1 23:09:25 2001 Brian Gough <bjg@network-theory.co.uk>
* gsl_matrix_float.h (struct ): MS VC++ requires that the
struct/typedef be made with a single definition
Mon Mar 26 20:33:45 2001 Brian Gough <bjg@network-theory.co.uk>
* view_source.c: split view functions out into a separate file
Sat Sep 9 16:47:16 2000 Brian Gough <bjg@network-theory.co.uk>
* added an owner field for indicating whether the underlying
memory is owned by the vector. Changed the meaning of the block
field to always be the address of the underlying block, even for
subviews (previously the block field was set to NULL in this
case).
Thu Aug 17 19:46:22 2000 Brian Gough <bjg@network-theory.co.uk>
* swap_source.c (FUNCTION): added function
gsl_matrix_transpose_memcpy for transposing rectangular matrices
Sun Jul 16 10:40:15 2000 Brian Gough <bjg@network-theory.co.uk>
* init_source.c (FUNCTION): added gsl_matrix_view function for
creating a matrix view of an ordinary C array
Sat Jul 15 21:44:22 2000 Brian Gough <bjg@network-theory.co.uk>
* changed GSL_EDOM to GSL_EINVAL for invalid matrix size arguments
Sat Jun 17 15:38:30 2000 Brian Gough <bjg@network-theory.co.uk>
* fixed up missing MULTIPLICITY factors in various functions
Sun May 28 12:25:02 2000 Brian Gough <bjg@network-theory.co.uk>
* test_complex_source.c (FUNCTION): use binary mode "b" when
reading and writing binary files
* test_source.c (FUNCTION): use binary mode "b" when reading and
writing binary files
Thu May 4 20:58:59 2000 Brian Gough <bjg@network-theory.co.uk>
* oper.c: added simple arithmetic operations (+,-,*,/,scale,+const)
Wed Apr 26 15:04:22 2000 Brian Gough <bjg@network-theory.co.uk>
* prop_source.c (FUNCTION): added function gsl_matrix_isnull
Tue Apr 18 12:51:49 2000 Brian Gough <bjg@network-theory.co.uk>
* minmax_source.c (FUNCTION): fixed bug in minmax where
coordinates would be incorrect for min or max in the first element
* test_source.c (FUNCTION): added tests for max/min functions
Thu Apr 13 18:39:27 2000 Brian Gough <bjg@network-theory.co.uk>
* minmax.c: added matrix max/min functions
Sat Mar 25 20:29:41 2000 Brian Gough <bjg@network-theory.co.uk>
* swap_source.c (FUNCTION): renamed swap_cols to swap_columns, and
renamed swap_rowcol to swap_row_column
Tue Mar 21 21:15:22 2000 Brian Gough <bjg@network-theory.co.uk>
* matrix_source.c (FUNCTION): added set_all and set_zero functions
Sat Mar 11 11:19:05 2000 Brian Gough <bjg@network-theory.co.uk>
* init_source.c (FUNCTION): added gsl_matrix_identity to set a
matrix to the identity, and gsl_matrix_zero for zeroing a matrix
* gsl_matrix.h: renamed struct element dim2 to tda (trailing
dimension) following blas conventions for row-major matrices
Thu Dec 2 21:17:16 1999 Brian Gough <bjg@network-theory.co.uk>
* rowcol_source.c (FUNCTION): added gsl_matrix_view_from_vector
(Thanks to Fabrice Rossi)
Sun Oct 31 20:01:39 1999 Brian Gough <bjg@network-theory.co.uk>
* copy.c copy_source.c: added gsl_matrix_copy function
Tue Oct 19 14:13:35 1999 Brian Gough <bjg@network-theory.co.uk>
* added gsl_matrix_swap_row/col to exchange rows and columns
Fri Oct 1 15:48:07 1999 Brian Gough <bjg@network-theory.co.uk>
* the matrix struct now supports a separate 'trailing dimension'
to allow handling of submatrices
* now uses separate block directory for memory management
Mon Mar 1 20:05:52 1999 Brian Gough <bjg@netsci.freeserve.co.uk>
* rowcol_source.c: fix rowcol to use strides
1998-11-09 <bjg@ancho.lanl.gov>
* test_source.c, test_complex_source.c: use macros to determine if
we should run tests with long double printf/scanf, since these
aren't supported on all platforms
Mon Apr 6 15:06:38 1998 Brian Gough <bjg@vvv.lanl.gov>
* make range checking the default, you have to define
GSL_RANGE_CHECK_OFF to turn it off