260bd57 | janv@uic.edu | 24 September 2015, 20:08:02 UTC | added "with Black_Polyhedral_Continuations" to phcpack.adb | 24 September 2015, 20:08:02 UTC |
57aa391 | janv@uic.edu | 24 September 2015, 18:39:14 UTC | adjusting READ_ME files for release 2.4.01 of PHCpack | 24 September 2015, 18:39:14 UTC |
f597ae3 | janv@uic.edu | 24 September 2015, 18:15:50 UTC | prepared for release 0.3.0 of phcpy | 24 September 2015, 18:15:50 UTC |
2bb5223 | janv@uic.edu | 24 September 2015, 15:51:54 UTC | updated greeting banners for release 2.4.01 of PHCpack | 24 September 2015, 15:51:54 UTC |
d0a3a68 | janv@uic.edu | 24 September 2015, 15:38:57 UTC | added blackbox solvers for Laurent polynomial systems in double double and quad double precision | 24 September 2015, 15:38:57 UTC |
7ac06f3 | janv@uic.edu | 24 September 2015, 15:10:05 UTC | black box root counters support double double and quad double precision | 24 September 2015, 15:10:05 UTC |
239ddcc | janv@uic.edu | 24 September 2015, 14:44:02 UTC | added push procedure for solutions with double double and quad double complex coefficients | 24 September 2015, 14:44:02 UTC |
086ac67 | janv@uic.edu | 23 September 2015, 14:00:36 UTC | extended version of the solvers of binomial systems in quad double precision in black box mode | 23 September 2015, 14:00:36 UTC |
2acaea2 | janv@uic.edu | 23 September 2015, 13:59:26 UTC | extended root refiners in quad double precision with reporting version | 23 September 2015, 13:59:26 UTC |
91afde9 | janv@uic.edu | 22 September 2015, 22:42:01 UTC | added functions to compute the size limit of string representations of polynomials, used in the Python interface to load polynomials from the system containers into a Python session | 22 September 2015, 22:42:01 UTC |
d658307 | janv@uic.edu | 22 September 2015, 20:32:26 UTC | changes in syscon caused changes in message passing routines | 22 September 2015, 20:32:26 UTC |
ac2aabf | janv@uic.edu | 22 September 2015, 20:04:12 UTC | more consistent naming of C functions that involve the standard systems containers | 22 September 2015, 20:04:12 UTC |
179cd48 | janv@uic.edu | 22 September 2015, 16:11:23 UTC | provided gateway functions to return the size limit of (Laurent) polynomials stored in the systems containers | 22 September 2015, 16:11:23 UTC |
da50cba | janv@uic.edu | 22 September 2015, 15:06:35 UTC | added function to compute an upper bound on the size of the string representations of polynomials | 22 September 2015, 15:06:35 UTC |
e6d4d9f | janv@uic.edu | 22 September 2015, 13:41:48 UTC | writing to strings of multivariate Laurent polynomials and multiprecision versions adjusted so stack_overflow is less likely to occur for huge polynomials | 22 September 2015, 13:41:48 UTC |
37c7c40 | janv@uic.edu | 21 September 2015, 22:26:26 UTC | increased the fixed limit on the number of characters in a string representation of a polynomial in several variables | 21 September 2015, 22:26:26 UTC |
de90dfe | janv@uic.edu | 21 September 2015, 22:25:39 UTC | updated makefiles for new test program | 21 September 2015, 22:25:39 UTC |
18b527f | janv@uic.edu | 21 September 2015, 22:22:20 UTC | the new recursive writing of a polynomial to a string makes fewer recursive calls and avoids stack_overflow with polynomial that have 700 hundred terms | 21 September 2015, 22:22:20 UTC |
24de59b | janv@uic.edu | 21 September 2015, 14:21:02 UTC | extended main black box solvers for one single equation to double double and quad double precision | 21 September 2015, 14:21:02 UTC |
7e8b2dd | janv@uic.edu | 21 September 2015, 14:18:26 UTC | extended the solvers of one polynomial equation in one variable in black box mode to double double and quad double precision | 21 September 2015, 14:18:26 UTC |
7daefc4 | janv@uic.edu | 20 September 2015, 20:47:39 UTC | adjusted procedure call for being in a new package | 20 September 2015, 20:47:39 UTC |
3ae500b | janv@uic.edu | 20 September 2015, 20:46:21 UTC | black box factorization runs in double double and quad double precision | 20 September 2015, 20:46:21 UTC |
b7833e2 | janv@uic.edu | 19 September 2015, 15:31:29 UTC | fixed dependency on a new package in a test program | 19 September 2015, 15:31:29 UTC |
12f80fe | janv@uic.edu | 19 September 2015, 11:00:34 UTC | fixed bug in normalize operation when expanding the trace form in double double and quad double precision of the interpolating polynomial | 19 September 2015, 11:00:34 UTC |
7008bc5 | janv@uic.edu | 19 September 2015, 09:42:10 UTC | removed code from the test program, code that was exported by the drivers | 19 September 2015, 09:42:10 UTC |
74d2e1a | janv@uic.edu | 18 September 2015, 14:11:47 UTC | updated drivers to call the factorization routines for multivariate polynomials with complex coefficients in double double and quad double precision | 18 September 2015, 14:11:47 UTC |
5ddd650 | janv@uic.edu | 18 September 2015, 09:57:32 UTC | extended factorization routines with monodromy operations in double double and quad double precision arithmetic | 18 September 2015, 09:57:32 UTC |
fe21147 | janv@uic.edu | 18 September 2015, 09:30:30 UTC | extended procedure to perform a monodromy breakup of multivariate polynomials with complex coefficients in double double and quad double precision | 18 September 2015, 09:30:30 UTC |
cf83c77 | janv@uic.edu | 18 September 2015, 08:24:08 UTC | added routines for the combinatorial exploration with the linear trace in double double and quad double precision to factor multivariate polynomials with complex coefficients | 18 September 2015, 08:24:08 UTC |
c8ce261 | janv@uic.edu | 17 September 2015, 13:59:08 UTC | new package to interpolate a factor given by witness points, with extracted code, and extended for double double and quad double precision | 17 September 2015, 13:59:08 UTC |
b28292f | janv@uic.edu | 17 September 2015, 12:49:55 UTC | new package to certify one factor of a single multivariate polynomial with the linear trace, with extracted code and extended for double double and quad double precision | 17 September 2015, 12:49:55 UTC |
15e9372 | janv@uic.edu | 17 September 2015, 09:00:52 UTC | new package with factored out coded and extended for double double and quad double precision | 17 September 2015, 09:00:52 UTC |
6261079 | janv@uic.edu | 16 September 2015, 10:20:21 UTC | new packages for divided differences in double double and quad double precision | 16 September 2015, 10:20:21 UTC |
76b87c0 | janv@uic.edu | 16 September 2015, 09:25:48 UTC | new packages for grids of sample points on hypersurfaces in double double and quad double precision | 16 September 2015, 09:25:48 UTC |
f8d5d7e | janv@uic.edu | 16 September 2015, 07:39:42 UTC | edits to make room for versions running in higher precision | 16 September 2015, 07:39:42 UTC |
e02faf1 | janv@uic.edu | 15 September 2015, 14:52:37 UTC | computing sequences of generic points on hypersurfaces in double double and quad double precision is done in new packages | 15 September 2015, 14:52:37 UTC |
9829388 | janv@uic.edu | 14 September 2015, 19:46:56 UTC | clarified dependencies for generic points computations | 14 September 2015, 19:46:56 UTC |
ac81ce0 | janv@uic.edu | 14 September 2015, 19:43:50 UTC | removed superfluous with/use cause pointed out by gnat gpl 2015 | 14 September 2015, 19:43:50 UTC |
12ea97a | janv@uic.edu | 14 September 2015, 19:41:17 UTC | updates because of new package standard_lined_hypersurfaces | 14 September 2015, 19:41:17 UTC |
5512482 | janv@uic.edu | 14 September 2015, 19:36:11 UTC | refactored code from hypersurface_samplers into the new standard_lined_hypersurfaces | 14 September 2015, 19:36:11 UTC |
53db748 | janv@uic.edu | 13 September 2015, 18:53:55 UTC | reformatting in the code dedicated to compute generic points on hypersurfaces to make room for the versions in double double and quad double precision | 13 September 2015, 18:53:55 UTC |
bff4351 | janv@uic.edu | 12 September 2015, 17:58:29 UTC | added procedures to deal with special case of solving linear systems in black box mode with doble double and quad double arithmetic | 12 September 2015, 17:58:29 UTC |
78e0fc5 | janv@uic.edu | 11 September 2015, 18:09:55 UTC | updates for double double and quad double precision black box solving of binomial and simplex systems | 11 September 2015, 18:09:55 UTC |
9aa85f8 | janv@uic.edu | 11 September 2015, 18:00:06 UTC | verbose versions of the root refiners in double double and quad double precision | 11 September 2015, 18:00:06 UTC |
9781ab3 | janv@uic.edu | 10 September 2015, 15:09:04 UTC | solving linear systems in double double and quad double precision when given as polynomial systems in symbolic format | 10 September 2015, 15:09:04 UTC |
2a01d61 | janv@uic.edu | 10 September 2015, 00:54:20 UTC | black box root counters contain polyhedral homotopies in quad double precision; the version string is printed at the end of the output file of the blackbox solver | 10 September 2015, 00:54:20 UTC |
fe19516 | janv@uic.edu | 09 September 2015, 21:59:48 UTC | blackbox root counters on Laurent systems can run polyhedral homotopies in double double precision | 09 September 2015, 21:59:48 UTC |
117c876 | janv@uic.edu | 09 September 2015, 12:07:48 UTC | path tracking for polyhedral homotopies with quad double precision arithmetic in blackbox mode | 09 September 2015, 12:07:48 UTC |
3791447 | janv@uic.edu | 09 September 2015, 12:00:03 UTC | added packages to make random coefficient systems with quad double precision coefficients | 09 September 2015, 12:00:03 UTC |
1de2ded | janv@uic.edu | 09 September 2015, 02:48:04 UTC | polyhedral continuation in blackbox mode with double double arithmetic | 09 September 2015, 02:48:04 UTC |
c73d0f9 | janv@uic.edu | 09 September 2015, 02:46:00 UTC | new package to randomize coefficients of polynomials in double double precision | 09 September 2015, 02:46:00 UTC |
9b8c53f | janv@uic.edu | 08 September 2015, 00:47:58 UTC | added with/use to package with refactored code from the blackbox root counters | 08 September 2015, 00:47:58 UTC |
340b842 | janv@uic.edu | 08 September 2015, 00:41:08 UTC | refactored code from the blackbox methods for polyhedral continuations | 08 September 2015, 00:41:08 UTC |
586c450 | janv@uic.edu | 06 September 2015, 21:53:33 UTC | multitasking of path trackers in black box mode enabled for double double and quad double precision | 06 September 2015, 21:53:33 UTC |
9deda76 | janv@uic.edu | 06 September 2015, 21:18:19 UTC | updated multitasked path trackers to Laurent homotopies in double double and quad double precision | 06 September 2015, 21:18:19 UTC |
8bdfdc2 | janv@uic.edu | 06 September 2015, 18:01:15 UTC | provided wrappers to the path trackers for Laurent systems in double double and quad double precision | 06 September 2015, 18:01:15 UTC |
893221b | janv@uic.edu | 06 September 2015, 01:36:47 UTC | running polynomial continuation with quad doubles in blackbox mode | 06 September 2015, 01:36:47 UTC |
12e2200 | janv@uic.edu | 06 September 2015, 01:31:37 UTC | Laurent and coefficient homotopies in quad double precision | 06 September 2015, 01:31:37 UTC |
2afb191 | janv@uic.edu | 06 September 2015, 01:14:26 UTC | updated makefile for new test program | 06 September 2015, 01:14:26 UTC |
84e7481 | janv@uic.edu | 06 September 2015, 01:08:34 UTC | new package to call path trackers in double double precision in blackbox mode | 06 September 2015, 01:08:34 UTC |
51bc361 | janv@uic.edu | 06 September 2015, 01:01:41 UTC | package for homotopies of Laurent systems with coefficients in double double precision | 06 September 2015, 01:01:41 UTC |
ee2c6d8 | janv@uic.edu | 04 September 2015, 15:32:13 UTC | reformatted code on the host for better readability | 04 September 2015, 15:32:13 UTC |
49ea35f | janv@uic.edu | 03 September 2015, 18:53:36 UTC | used the verbose option when calling Newton's method on the GPU | 03 September 2015, 18:53:36 UTC |
39d6de6 | janv@uic.edu | 03 September 2015, 18:50:16 UTC | printing of residual is guarded by the if(debug) statement in GPU_Newton | 03 September 2015, 18:50:16 UTC |
406ebcd | janv@uic.edu | 03 September 2015, 18:10:59 UTC | added verbose option to Newton's method on the GPU | 03 September 2015, 18:10:59 UTC |
949aaf1 | janv@uic.edu | 03 September 2015, 16:26:49 UTC | documented the values for the monomial evaluation method | 03 September 2015, 16:26:49 UTC |
4a6c777 | janv@uic.edu | 03 September 2015, 00:54:34 UTC | updated CUDA code for Newton's method and path tracking | 03 September 2015, 00:54:34 UTC |
55adad1 | janv@uic.edu | 03 September 2015, 00:50:55 UTC | updated code for evaluation and Newton's method on the host | 03 September 2015, 00:50:55 UTC |
d387068 | janv@uic.edu | 03 September 2015, 00:22:00 UTC | fixed MON_EVAL_METHOD to 1 as needed to track one path with quad doubles | 03 September 2015, 00:22:00 UTC |
be9a2b5 | janv@uic.edu | 01 September 2015, 22:03:02 UTC | updated makefile to make the test program ts_mva | 01 September 2015, 22:03:02 UTC |
8d72888 | janv@uic.edu | 01 September 2015, 22:02:22 UTC | added callback procedure to the mixedvol algorithm | 01 September 2015, 22:02:22 UTC |
c1f04bb | janv@uic.edu | 31 August 2015, 22:37:06 UTC | release 2.4 of PHCpack is still marked as 2.4.00 | 31 August 2015, 22:37:06 UTC |
ca75c02 | janv@uic.edu | 31 August 2015, 22:22:33 UTC | version 0.2.9 of phcpy is ready for release | 31 August 2015, 22:22:33 UTC |
8d148c6 | janv@uic.edu | 31 August 2015, 21:52:38 UTC | added script to call the GPU accelerated path trackers in phcpy | 31 August 2015, 21:52:38 UTC |
1aad2cc | janv@uic.edu | 31 August 2015, 21:42:03 UTC | updated documentation with a speedup calculation of the wall clock times | 31 August 2015, 21:42:03 UTC |
9fb169c | janv@uic.edu | 31 August 2015, 21:16:37 UTC | the Ada code is ready for release 2.4 | 31 August 2015, 21:16:37 UTC |
d286cb4 | janv@uic.edu | 31 August 2015, 19:34:29 UTC | fixed calls to include values for the gamma constant | 31 August 2015, 19:34:29 UTC |
aa10691 | janv@uic.edu | 31 August 2015, 18:56:40 UTC | fixed bug in capturing number of tasks as passed from the C interface | 31 August 2015, 18:56:40 UTC |
b44d25a | janv@uic.edu | 31 August 2015, 00:22:49 UTC | updated makefiles for the new test programs and to make the new .so modules | 31 August 2015, 00:22:49 UTC |
026c64a | janv@uic.edu | 31 August 2015, 00:21:18 UTC | versions for double double and quad double precision to call the GPU accelerated path trackers | 31 August 2015, 00:21:18 UTC |
b288a1f | janv@uic.edu | 30 August 2015, 23:36:11 UTC | interface module to the double double and quad double versions of the GPU accelerated path trackers is done | 30 August 2015, 23:36:11 UTC |
0bcab7b | janv@uic.edu | 30 August 2015, 22:17:22 UTC | separate header files for the GPU acceleration of Newton's method | 30 August 2015, 22:17:22 UTC |
01acba7 | janv@uic.edu | 30 August 2015, 21:24:51 UTC | gpu_double_track added to trackers.py for GPU accelerated path tracking via Python | 30 August 2015, 21:24:51 UTC |
ae4e8ba | janv@uic.edu | 30 August 2015, 21:21:45 UTC | the interface C module is extended with GPU accelerated Newton and path trackers | 30 August 2015, 21:21:45 UTC |
367cb7d | janv@uic.edu | 30 August 2015, 19:58:28 UTC | added gamma constant to path trackers with algorithmic differentiation in double precision | 30 August 2015, 19:58:28 UTC |
106220f | janv@uic.edu | 30 August 2015, 19:07:03 UTC | applied extra input arguments for the random gamma constant to the interface C library routines | 30 August 2015, 19:07:03 UTC |
4565a07 | janv@uic.edu | 29 August 2015, 16:25:37 UTC | updated ad hoc interface to track one path on the GPU in quad double precision | 29 August 2015, 16:25:37 UTC |
4dd6846 | janv@uic.edu | 29 August 2015, 00:24:43 UTC | extended interface functions to the GPU path trackers with gamma constant | 29 August 2015, 00:24:43 UTC |
41bc1ee | janv@uic.edu | 28 August 2015, 16:12:36 UTC | interface to C functions for path tracking using algorithmic differentiation accept gamma constant as input parameter | 28 August 2015, 16:12:36 UTC |
d362e6d | janv@uic.edu | 27 August 2015, 12:48:59 UTC | bug fix in standard_double_track and new function to apply algorithmic differentiation in the path trackers in double precision | 27 August 2015, 12:48:59 UTC |
64ca49e | janv@uic.edu | 26 August 2015, 17:39:51 UTC | new makefiles for the extended Python interface module | 26 August 2015, 17:39:51 UTC |
3bf2955 | janv@uic.edu | 26 August 2015, 17:38:13 UTC | extension of phcpy2c interface module to Python with Newton's method and path trackers with algorithmic differentiation in double precision | 26 August 2015, 17:38:13 UTC |
de57964 | janv@uic.edu | 25 August 2015, 21:26:17 UTC | updated makefiles for new test programs | 25 August 2015, 21:26:17 UTC |
9982bab | janv@uic.edu | 25 August 2015, 21:05:13 UTC | C interface in PHClib to the functions that apply algorithmic differentiation in Newton's method and the path trackers in double, double double, and quad double precision | 25 August 2015, 21:05:13 UTC |
8f7d49c | janv@uic.edu | 25 August 2015, 20:32:40 UTC | separated prototypes of functions into header files for the application of algorithmic differentiation in Newton's method and the path trackers | 25 August 2015, 20:32:40 UTC |
2ff7592 | janv@uic.edu | 24 August 2015, 21:56:05 UTC | updated makefiles for the new test programs | 24 August 2015, 21:56:05 UTC |
0c02f70 | janv@uic.edu | 24 August 2015, 21:54:27 UTC | quad double versions for the wrappers to execute Newton's method and the path trackers with algorithmic differentiation to evaluate the polynomials | 24 August 2015, 21:54:27 UTC |
47add75 | janv@uic.edu | 24 August 2015, 21:35:25 UTC | still minor error that occurs when the double double code is verbose | 24 August 2015, 21:35:25 UTC |
a53c340 | janv@uic.edu | 24 August 2015, 21:28:14 UTC | adjusted function calls to PHClib for double double versions of copying target, start system, and retrieving the number of solutions | 24 August 2015, 21:28:14 UTC |
e53fe4a | janv@uic.edu | 23 August 2015, 17:47:14 UTC | updated makefiles for new test programs | 23 August 2015, 17:47:14 UTC |