|
Packit |
67cb25 |
2018-03-18 Tuomo Keskitalo <tuomo.keskitalo@iki.fi>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* rk2imp.c and evolve.c: Added debug printouts, run indent -gnu -nut.
|
|
Packit |
67cb25 |
Thanks for Dominic Steinitz for the additions!
|
|
Packit |
67cb25 |
* test.c: Added identifiers to debug printouts.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2017-11-04 Tuomo Keskitalo <tuomo.keskitalo@iki.fi>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* cscal.c: Corrected bug #52336 in cscal.c: Decreased maximum step
|
|
Packit |
67cb25 |
length increase coefficient (now named maxscale) from 5.0 to 4.9
|
|
Packit |
67cb25 |
because value of 5.0 hits a singularity in msbdf stepper with
|
|
Packit |
67cb25 |
order 3, causing a floating point exception. Thanks to Andrew
|
|
Packit |
67cb25 |
Benson for reporting this bug!
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2017-10-22 Tuomo Keskitalo <tuomo.keskitalo@iki.fi>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* msadams.c: Corrected bug #52230 in msadams_apply, where solver
|
|
Packit |
67cb25 |
crashed to order change -2 error. This could happen in a very rare
|
|
Packit |
67cb25 |
case when two consecutive failed steps suggest decreasing
|
|
Packit |
67cb25 |
order. Fix is to reset solver if that happens. Thanks for Michael
|
|
Packit |
67cb25 |
Kaufman for reporting this bug and testing the fix!
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2013-06-08 Tuomo Keskitalo <tuomo.keskitalo@iki.fi>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* driver.c: Removed redundant if clause. Thanks to
|
|
Packit |
67cb25 |
David Binderman for the report!
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2013-01-27 Tuomo Keskitalo <tuomo.keskitalo@iki.fi>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* msbdf.c: Corrected bug which enabled order to be changed
|
|
Packit |
67cb25 |
by two (first by stability enhancement, then by order evaluation
|
|
Packit |
67cb25 |
after a rejected step). Thanks for Frantisek Kluknavsky and
|
|
Packit |
67cb25 |
Andreas Schwab for bug reports!
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* msbdf.c: Added more state variables explicitly to be reset in
|
|
Packit |
67cb25 |
msbdf_reset.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
*msbdf.c: Added abscorscaled to remove division of abscor for
|
|
Packit |
67cb25 |
use in msbdf_eval_order and subsequent backwards multiplication.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
*test.c: test_extreme_problems: Increased ringmod case tolerances
|
|
Packit |
67cb25 |
from 1e-7 to 1e-12 to increase precision. Because that
|
|
Packit |
67cb25 |
increases computational burden, I also decreased end time
|
|
Packit |
67cb25 |
from 1e-3 to 1e-5. That decreased the acidity of the test
|
|
Packit |
67cb25 |
significantly, but the test case is now more appropriate
|
|
Packit |
67cb25 |
for normal "make check". Thanks for Frantisek Kluknavsky
|
|
Packit |
67cb25 |
for pointing out bad design of the test case.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2012-09-10 Rhys Ulerich <rhys.ulerich@gmail.com>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* test.c: Correct two out-of-order declarations.
|
|
Packit |
67cb25 |
Thanks to Brian Gladman for spotting the problem.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2012-03-10 Tuomo Keskitalo <tuomo.keskitalo@iki.fi>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* driver.c: Added function gsl_odeiv2_driver_reset_hstart to
|
|
Packit |
67cb25 |
allow resetting step size, possibly change its sign, too.
|
|
Packit |
67cb25 |
Check for non-zero hstart in initializations.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* test.c: Modified test_driver to carry out tests with all
|
|
Packit |
67cb25 |
steppers. Small printout changes.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2012-01-21 Tuomo Keskitalo <tuomo.keskitalo@iki.fi>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* rkf45.c: Bug correction: Set gives_exact_dydt_out to 1 in
|
|
Packit |
67cb25 |
rkf45_type
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2012-01-14 Tuomo Keskitalo <tuomo.keskitalo@iki.fi>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* evolve.c: Modified initial derivative evaluation in evolve_apply
|
|
Packit |
67cb25 |
to reuse previously calculated values. This saves calls to the
|
|
Packit |
67cb25 |
user function. Thanks for Illes Farkas for pointing out redundant
|
|
Packit |
67cb25 |
function calls!
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2011-06-28 Brian Gough <bjg@network-theory.co.uk>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* rk4imp.c (rk4imp_apply): use M_SQRT3 instead of sqrt(3) in array
|
|
Packit |
67cb25 |
initialiser.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2011-05-02 Brian Gough <bjg@network-theory.co.uk>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* gsl_odeiv2.h: fix header guard to __GSL_ODEIV2_H__
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2011-05-01 Tuomo Keskitalo <Tuomo.Keskitalo@iki.fi>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* Replaced ChangeLog with contents of ode-initval2 development log
|
|
Packit |
67cb25 |
file (ChangeLog-odeiv2_development) and reformatted the log file.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2010-10-16 Tuomo Keskitalo <Tuomo.Keskitalo@iki.fi>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* Modified evolve and driver so that when user function returns
|
|
Packit |
67cb25 |
GSL_EBADFUNC, then ode-initval2 functions return immediately with
|
|
Packit |
67cb25 |
that same return value.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* Corrected bug in msadams for saving ordm1coeff to state. Thanks
|
|
Packit |
67cb25 |
to Andrew Benson for pointing out this bug!
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2010-10-10 Tuomo Keskitalo <Tuomo.Keskitalo@iki.fi>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* driver.c: Corrected error in driver_evolve_apply, integration to
|
|
Packit |
67cb25 |
negative direction should work now.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* Added evolve_apply_fixed_step and driver_apply_fixed_step for
|
|
Packit |
67cb25 |
users who want to use a constant step size to evolve the system.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* test.c: added test cases for driver (negative evolution,
|
|
Packit |
67cb25 |
apply_fixed_step).
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* gsl_odeiv2.h: cleaning and add apply_fixed_step functions.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* doc/ode-initval.texi: Replaced derivative apostrophe with \prime
|
|
Packit |
67cb25 |
(because my new system fails to make ps otherwise). Added new
|
|
Packit |
67cb25 |
documentation about apply_fixed_step functions.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2010-10-09 Tuomo Keskitalo <Tuomo.Keskitalo@iki.fi>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* ode-initval2 bzr development branch now at savannah:
|
|
Packit |
67cb25 |
http://bzr.savannah.gnu.org/r/gsl/ode-initval2/
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2010-01-02 Tuomo Keskitalo <tuomo.keskitalo@iki.fi>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* Changes from ode-initval2-0.9 to ode-initval2-1.0:
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* Removed set_control member from step object and added a
|
|
Packit |
67cb25 |
set_driver member to evolve, control and step objects, so that the
|
|
Packit |
67cb25 |
driver object and consequently evolve, control and step objects
|
|
Packit |
67cb25 |
are all accessible from another. Currently only some steppers
|
|
Packit |
67cb25 |
utilize this. They access the control_errlevel function via
|
|
Packit |
67cb25 |
state->driver->c instead of state->control. This modification to
|
|
Packit |
67cb25 |
the framework enables implementation of more efficient ODE
|
|
Packit |
67cb25 |
solvers, which utilize communication among evolve, control and
|
|
Packit |
67cb25 |
stepper objects. This also means that some steppers now require a
|
|
Packit |
67cb25 |
driver object, and it is best to use those stepper methods through
|
|
Packit |
67cb25 |
driver object functions.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* cstd.c and cscal.c: Added a sanity check for errlev <= 0
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* run indent -gnu -nut for *.c *.h
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* driver.c: added function odeiv2_driver_alloc_standard_new and
|
|
Packit |
67cb25 |
odeiv2_driver_alloc_yp_new
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* driver.c: changed n and nmax in driver struct to unsigned long
|
|
Packit |
67cb25 |
int
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* modified documentation accordingly with small nomenclature
|
|
Packit |
67cb25 |
changes and made year updates to source files
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2009-11-01 Tuomo Keskitalo <tuomo.keskitalo@iki.fi>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* Merged with master and modified source to be published as a
|
|
Packit |
67cb25 |
package for gsl-1.13.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2009-05-09 Tuomo Keskitalo <tuomo.keskitalo@iki.fi>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* Agreed with Brian Gough that the modification I've made cannot
|
|
Packit |
67cb25 |
be incorporated to GSL until a major version number change, due to
|
|
Packit |
67cb25 |
changes in API.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* Replated rk2imp with imprk2 and rk4imp with imprk4. Also renamed
|
|
Packit |
67cb25 |
impeuler to rk1imp for consistency.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* Deleted steppers that are deprecated considering new major
|
|
Packit |
67cb25 |
release: gear1, gear2, and rk2simp. Modified test.c accordingly.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2009-05-01 Tuomo Keskitalo <tuomo.keskitalo@iki.fi>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* Merged with master (evolve.c and test.c needed some manual
|
|
Packit |
67cb25 |
work).
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* driver.c: Modified alloc-functions to minimal interface, added
|
|
Packit |
67cb25 |
set_hmax, set_nmax and set_hmin functions.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* test.c: Added test for minimum allowed step size.
|
|
Packit |
67cb25 |
Applied indent -gnu -nut.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* doc/ode-initval.texi: Updated documentation, added Driver
|
|
Packit |
67cb25 |
section, updated examples.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2009-04-05 Tuomo Keskitalo <tuomo.keskitalo@iki.fi>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* Several uninitialized variable corrections from Brian Gough
|
|
Packit |
67cb25 |
added
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* modnewton1, impeuler, imprk2, imprk4: removed unneeded arguments
|
|
Packit |
67cb25 |
from solve and init
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* msadams and msbdf modifications:
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* added ordprevbackup, ordwaitbackup (backups) and failcount (to
|
|
Packit |
67cb25 |
detect repeated steps rejections).
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* added checks to make sure order is not changed by more than one
|
|
Packit |
67cb25 |
order. Added sanity check to catch it.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* corrected failure and rejection handling for first step
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* Added a heuristic stability enhancement for msbdf. This is yet
|
|
Packit |
67cb25 |
to be tested in practice.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* test.c: Added more benchmark test cases. The results seem
|
|
Packit |
67cb25 |
sensible.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* test.c: Note for benchmark_precision: msbdf performs badly in
|
|
Packit |
67cb25 |
rhs_func_exp. Newton iteration starts to fail to converge at some
|
|
Packit |
67cb25 |
point, and this forces very small step sizes to be used. I
|
|
Packit |
67cb25 |
currently think that this is not a bug, but that msbdf is just not
|
|
Packit |
67cb25 |
suitable for this problem. I think that msbdf's prediction step
|
|
Packit |
67cb25 |
does not give a good initial guess for Newton iteration in this
|
|
Packit |
67cb25 |
case.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* I tried to increase max_iter in msbdf from 3 to 7 (value used in
|
|
Packit |
67cb25 |
modnewton1), and as a result, the performance was better for this
|
|
Packit |
67cb25 |
case (Newton iteration convergence was better). However, I would
|
|
Packit |
67cb25 |
not increase max_iter, because the default value of 3 is the value
|
|
Packit |
67cb25 |
used by the authors in the references, and because it would
|
|
Packit |
67cb25 |
decrease the efficiency of msbdf (method would use more function
|
|
Packit |
67cb25 |
evaluations per step, and convergence might still not be
|
|
Packit |
67cb25 |
achieved). It is best to useanother stepper in this kind of a
|
|
Packit |
67cb25 |
case.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2009-02-28 Tuomo Keskitalo <tuomo.keskitalo@iki.fi>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* new stepper: msbdf. This is a variable-coefficient linear
|
|
Packit |
67cb25 |
multistep backward differentiation formula (BDF) method in
|
|
Packit |
67cb25 |
Nordsieck form. It appears to handle stiff cases well and it is
|
|
Packit |
67cb25 |
efficient!
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* test.c: Ordered functions in file. Added extreme test case.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* modnewton1: Convergence test now takes the desired error level
|
|
Packit |
67cb25 |
via control object into account. Modified steppers that use
|
|
Packit |
67cb25 |
modnewton1 accordingly.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* modnewton1: Removed reuse_decomp. The variable was useless since
|
|
Packit |
67cb25 |
iteration matrix is reconstructed at each stepper call.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* modnewton1: Dropped max_iter from 10 to 7. Hairer and Wanner
|
|
Packit |
67cb25 |
suggests a "relatively high number", but many iterations per step
|
|
Packit |
67cb25 |
is costly.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* steppers that use step doubling for error estimates: Introduced
|
|
Packit |
67cb25 |
constant GSL_ODEIV_ERR_SAFETY in gsl_odeiv.h that is used to
|
|
Packit |
67cb25 |
multiply the error estimates (yerr). This way user can modify this
|
|
Packit |
67cb25 |
coefficient in one place if a change is wanted.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* driver.c: This is a high level wrapper for ode-initval, which
|
|
Packit |
67cb25 |
was wished for in TODO. Added test_driver to test.c to test it. Is
|
|
Packit |
67cb25 |
this functionality enough?
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* ode-initval.texi: Added documentation for new steppers. Moved
|
|
Packit |
67cb25 |
note about user error codes to system definition.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* Some finetuning / code cleanup.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2009-01-21 Tuomo Keskitalo <tuomo.keskitalo@iki.fi>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* New explicit stepper: msadams (An explicit variable-coefficient
|
|
Packit |
67cb25 |
linear multistep Adams method in Nordsieck form). Functional
|
|
Packit |
67cb25 |
iteration is used to solve the non-linear equation system. The
|
|
Packit |
67cb25 |
algorithms described in the references (see msadams.c) have been
|
|
Packit |
67cb25 |
adapted to fit GSL odeiv framework. Currently this is beta, but
|
|
Packit |
67cb25 |
the stepper now passes the odeiv test suite.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* control.c, cstd.c, cscal.c: Added errlev function which
|
|
Packit |
67cb25 |
calculates desired error level. This is used in masadams (and can
|
|
Packit |
67cb25 |
be used in imp* steppers, too) to control accuracy of functional
|
|
Packit |
67cb25 |
iteration. User must call gsl_odeiv_step_set_control after
|
|
Packit |
67cb25 |
allocating step and control objects to pass the information about
|
|
Packit |
67cb25 |
control object to stepper. This change (or something similar) is
|
|
Packit |
67cb25 |
needed to pass tolerance for the stepper routines that need
|
|
Packit |
67cb25 |
it. Modified stepper routines accordingly.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* evolve.c: Changed the step length scaling coefficient for
|
|
Packit |
67cb25 |
stepper failure from 0.1 to 0.5, because msadams seems to cope
|
|
Packit |
67cb25 |
much better with this change in stiff[0,1] test case.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* gsl_odeiv.h: gsl_odeiv_control and control_type introduced
|
|
Packit |
67cb25 |
earlier because gsl_odeiv_step_type now refers to it.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
2008-11-01 Tuomo Keskitalo <tuomo.keskitalo@iki.fi>
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* New implicit solvers: imprk4 (replaces imprk42 in previous test
|
|
Packit |
67cb25 |
package), imprk2 (replaces imprk21 in previous test package) and
|
|
Packit |
67cb25 |
impeuler, which use modified Newton iteration to solve the
|
|
Packit |
67cb25 |
implicit equations. I decided to dump imprk42 and imprk21 because
|
|
Packit |
67cb25 |
I was unsure how to calculate the error estimates for them. I
|
|
Packit |
67cb25 |
think that these new implicit methods could be used instead of
|
|
Packit |
67cb25 |
rk4imp, rk2imp, gear2 and gear1, which use functional
|
|
Packit |
67cb25 |
iteration (which is inefficient for stiff problems.) However, even
|
|
Packit |
67cb25 |
these new solvers can not be used efficiently for truly nasty
|
|
Packit |
67cb25 |
stiff problems, and better solvers are still needed.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* modnewton1.c: a modified Newton solver for solution of
|
|
Packit |
67cb25 |
non-linear equations of implicit solvers. Jacobian evaluation has
|
|
Packit |
67cb25 |
been moved from modnewton1 to imprk2, imprk4 and impeuler.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* evolve.c: modification to decrease step size if stepper fails,
|
|
Packit |
67cb25 |
instead of giving up immediately.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* evolve.c: A bug fix: Exit with GSL_FAILURE if step size reaches
|
|
Packit |
67cb25 |
machine precision instead of continuing with old step size.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* test.c: added Robertson stiff test
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* test.c: embedded test_oregonator to test_compare_stiff_probelms
|
|
Packit |
67cb25 |
to test several stiff problems. Note: the tested integration
|
|
Packit |
67cb25 |
interval is rather short due to inefficiency of solvers other than
|
|
Packit |
67cb25 |
bsimp in these problems.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* test.c: modified test_compare_vanderpol and
|
|
Packit |
67cb25 |
test_compare_stiff_problems to check against results from first
|
|
Packit |
67cb25 |
ode-solver. Changed rk4 or bsimp to first place.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* test.c: modified stepfn and stepfn2 and expanded these to be
|
|
Packit |
67cb25 |
tested with all explicit solvers
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* test.c: removed stepfn3 and added test_broken and
|
|
Packit |
67cb25 |
test_stepsize_fail to check that evolve decreases step size below
|
|
Packit |
67cb25 |
machine precision and exits with a failure code in the end if user
|
|
Packit |
67cb25 |
functions fail continuously.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* Changes in ode-initval.texi:
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* removed untrue sentence concerning stepping functions: "The
|
|
Packit |
67cb25 |
step-size @var{h} will be set to the step-size which caused the
|
|
Packit |
67cb25 |
error."
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* Changed gsl_odeiv_system variable name from dydt to sys for
|
|
Packit |
67cb25 |
clarity
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* added "explicit" or "implicit" to stepper introductions for
|
|
Packit |
67cb25 |
clarity
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* reformulated description of evolve_apply and step_apply and made
|
|
Packit |
67cb25 |
changes to reflect the modifications to the code.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* clarified the point on integrating over discontinuities. It is
|
|
Packit |
67cb25 |
best to integrate in sequences.
|
|
Packit |
67cb25 |
|
|
Packit |
67cb25 |
* removed the text suggesting the user to force a step size to
|
|
Packit |
67cb25 |
integrate over the edge of a discontinuity. The applicability of
|
|
Packit |
67cb25 |
this kind of a numerical tweak depends on the case, and should in
|
|
Packit |
67cb25 |
my opinion not to be included in a reference book.
|