NEWS
optimx 2024-12-31
- Change positive-definitive check on Hessian to pd_check(). This is in
routine kktchk which calls pd_check(). Note that we accept the default
arguments pivot=TRUE and tol=1.0e-7 for that routine. Do we want to allow
users to supply these?
- Added a new control to ctrldefault() called kkt2chol, default TRUE, which
uses a Cholesky decomposition attempt as the test. FALSE uses eigenvalue
ratio test. See kktchk() code for details.
optimx 2024-12-18
- Preserve parameter names in optimr()
- Warnings about with() in opm.Rd and optimr.Rd
optimx 2024-12-02
- Fix gradient count in scounts for approximations
- Limit proptimr to nlim output items in vectors and Hessian rows and cols.
- Fix reset of parameters in grfwd
- Minor fix to Rtnmin.Rd
- Remove extra material from /inst/doc to reduce size
- Ensure maxfeval gets set correctly in ncg
- Rationalize Rd files for Rvmmin, Rcgmin, Rtnmin
optimx 2024-10-22
- Note that stopping when number of function, gradient or hessian evaluations
reached is difficult as solvers have many different approaches, and some
lack such controls. Note that many talk of "iterations". We will use maxit
to be the limit on gradient evaluations. Users could adjust the efn(), egr()
and ehess() in function optimr() to force a halt if desired. Commented out
example code for optsp$kfn and optsp$kgr tests are in optimr.R.
- Fix tn() and tnbc() to take control list
- Incorporate maxfeval into Rcgmin, Rcgminb, Rcgminu
- Cleanup of a number of small issues.
optimx 2024-10-09
- Added vignette which shows addition of pracmanm() solver to optimr()
- Make Rtnmin the primary name for tn and tnbc tools (Truncated Newton)
optimx 2024-10-07
- checkallsolvers() offers to install missing packages (though only CRAN
packages currently)
optimx 2024-10-02
- Fixed error in description of grpracma Rd file.
- Github README.md added
optimx 2024-03-18
- Check that ans$convergence is defined for lbfgsb3c in optimr(). If NA,
then set ans$convergence to 9995
- opm(): remove non-mask methods from method list when masks requested
or upper and lower bounds coincide
optimx 2024-03-15
- Put counters in efn, egr, ehess. Counters are optsp$kfn, optsp$kgr,
and optsp$khe in the special environment optsp created by zzz.R.
These are NOT output by all methods.
Some comments apply to deprecated packages optimr on CRAN, github.com/nashjc/optimx
and optimrx on R-forge.
optimx 2023-10-27
- ncg / Rcgminu / Rcgminb -- typo in keepgoing (was keekpgoing) fixed.
- added opm2optimr and manual file for extracting single solver result
from opm() output object
optimx 2023-10-21
- Expand diagnostic for missing or unsuitable method(s)
- Fix unnecessary ",..." in call to L-BFGS-B in optimr(). The dot-args were
removed to avoid an argument collision (VERSION 2023-06-25) using a suggestion
of Duncan Murdoch, but I forgot to remove this case.
- Duncan Murdoch proposed and suggested a patch to avoid termination by try(),
though failure is reported, and to attempt to give better error messages on
cause of failure in such cases.
- Some minor corrections to simple typos and extraneous text.
- Added checks on valid methods in opm(), trap for bounded methods in optimr().
optimx 2023-09-01
- Used Henrik Bengsston's R.rsp method to install static pdf vignettes as some
CRAN test platforms seem to be missing files needed to build vignettes from
Rmarkdown source files. The vignette source Rmarkdown files are in the
directory inst/doc/vignette-source.
optimx 2023-08-23
- Thanks to Mark Pletscher for uncovering a bug due to a typo in optimr(). Also
noting that optimr() now fails if input parameters are infeasible. However,
users can pre-adjust parameters as follows
# set initial parameters in vector init, lower and upper bounds in lo and up
chkpar <- bmchk(init, lower=lo, upper=up)
if (chkpar$parchanged) init <- chckpar$bvec
- kktchk now has control with default dowarn=TRUE to allow suppression of overly
common warning that estimated Hessian at solution is asymmetric. The test for
asymmetry is very difficult to set up in a way that does not give warnings when
the Hessian estimate is reasonable (false positives).
optimx 2023-06-25
- This is a MAJOR revision and overhaul of the optimx package and its components.
- offset removed as control. Use reltest.
- gHgen.R and gHgenb.R are no longer needed by optimx and are kept purely for
legacy uses by others.
- optimr now tries to add function name attribute to returned 'value' object;
opm tries to add this attribute to the answer object.
- 20230730: added pracma::nelder_mead and nloptr::neldermead
- ctrldefault() now offers collections (vectors) of methods nogrmeth, grmeth, hessmeth
for those with no gradient requirement, gradient methods and methods that can or
must have Hessian function supplied
- opm() does more checks e.g., missing gr, methods inappropriate to problem
- nvm and Rvmminb: fixed issue that sometimes trystep (distance to bound
in search direction) came back as small NEGATIVE number) by setting
any steplength < relstep*.Machine$double.eps to 0. This forces a search
in gradient (steepest descent) direction and seems to correct runaway
backtracking that fails out on too much work.
- Fixed CITATION file based on R CMD check --as-cran complaints
regarding requirement for person() and bibentry() changes.
- In inst/doc/examples/fhop.R: found we need to ensure both up and lo
defined to get certain solvers to work e.g., lbfgsb3c. This meant
adjusting fhop to put Inf or -Inf in as appropriate.
- checksolver() and checkallsolvers() revised and documented. No
check now in ctrldefault() for solver availability. That is, we
will tolerate a failure and rerun if needed solvers are missing.
- Added "Intro to optimx" vignette. Specify that gradient function
should be supplied explicitly if possible.
- Add method="MOST" options to try most methods in opm(). This is
a list returned as item 'mostmeth' from ctrldefault().
- optimx legacy function fns use numDeriv explicitly for grad, jacobian,
and hessian
- dot arguments added in optimr() to subplex
- Importance of named arguments over positional ones noted in Intro vignette.
- Minor fix to test of maximize in optimx.setup()
- Legacy optimx() demos in inst/doc/legacy-demo/. Note that these have been
checked to ensure they "run", though results have NOT been verified.
- Examples in inst/doc/examples/ have all been checked to ensure they run.
Results have NOT all been verified in detail.
- A test of 'hessian="approx"' is in vignettes/GenRoseHess.R
- grchk() is now called in optimr, NOT solvers, except (for the moment)
in Rvmmin().
- Example script fhop.R that uses the funconstrain package tests now has
a check that funconstrain package is installed, and suggests how to
install from Github since it is not on CRAN.
- A long-standing issue is that approximated Hessian at return is NOT
symmetric. i.e., kktchk -- Symmetry check often fails and we symmetrize
the computed Hessian. See testing script inst/doc/examples/HessQuality.R
- demo() has been updated to use newer functions, but is not yet tidy.
- grpracma.R allows "grpracma" as gradient function and uses pracma approx.
ctrldefault() hesspkg and jacpkg controls to use pracma rather than
numDeriv for hessian and jacobian approximations.
getFromNamespace("fn","pkgname") to set up functions like jacobian, hessian
- Checked special controls and adjusted for bobyqa method. Note
inst/doc/examples/specctrlhobbs.R
- Added scounts to optimr, updated proptimr() to use these if available,
also summary.opm()
- Using ideas from optim() (thanks to Duncan Murdoch for pointer), dotargs
suppressed internally in fn, gr, hess to avoid argument clash as in
https://stackoverflow.com/questions/69033754/maximum-likelihood-estimation-of-three-parameter-reverse-weibull-model-implement
Added in optimx() as well as optimr()
- ncg reimplemented from Rcgmin to use checking done at start of optimr and
avoid duplication of effort. ncg() should NOT be used alone.
- Maximization problems have function value reversed in optimr() NOT opm()
- Add runex.R for running examples in inst/doc/examples/
- Stripped down nvm() to use checks already done in optimr()
- Commented out user input after error in line search in Rtnmin-package.R
- Fixed minor glitch in bmchk() status indicators
- Created first try at snewtm() to run only from optimr; snewtonm() left as
stand-alone function
optimx 2022-11-25
- kktchk -- Symmetry check often fails and we symmetrize the computed Hessian.
Note the testing script HessQuality.R (20220610)
- Rcgmin.Rd has examples changed to properly handle approximations via
optimr().
- Added note to opm.Rd that maxfeval is passed through to methods that can
use it.
- snewtonmb --> snewtonm while snewtonm -> snewtonmu (unconstrained)
as bounded verson faster with unbounded cases. Removed snewtonmu.R to
00unversioned. Script snewtonmbtest.R has an "all masked" test which works.
bmchk() modified so it correctly adjusts masks and only sets parchanged
when they parameters are adjusted
- optimr() returns a "status" attribute for bounds which proptimr() displays
if available
- control checkgrad is used to invoke function grchk() to check supposedly
correct (analytic) gradient code.
- Change bmchk to use an offset for checking equality -- otherwise get poor
results from "status" indicator on parameters
- The Argonne tests (More', J. J., Garbow, B. S., & Hillstrom, K. E. (1981).
Testing unconstrained optimization software. ACM Transactions on Mathematical
Software (TOMS)}, {7}(1), 17-41. doi.org/10.1145/355934.355936) in the
funconstrain package to which J C Nash added Hessians,
https://github.com/jlmelville/funconstrain
Now have a convenient test script fhop.R
- Replaced doi in CITATION because R CMD check complained. Possibly JSS has
changed some specs on its doi values.
- Added 'weakmeth' to ctrldefault() and these methods do NOT contribute to
opm() with method="ALL"
- Many warnings 'pHes not symmetric' in ffn runs. Tried to scale to nHes, but
unsymmetric hessian gives complex eigenvalues, so NOT a good idea to ignore.
- added marqLevAlg::mla to optimr. Note extremely different return structure
from optim() etc.
- ncg and ncgqs issue with bdstest with masks --
needed to pass bdmsk to solver correctly in optimr
- returned hessian uses hess() if available, else approx.
- opm and summary.opm add bounds status.
- "probname" attribute to ans$value added to output of optimr.
- Status indicators for parameters (F, L, U) in optimr, proptimr.
- proptimr() displays bounds, masks info from bmchk$bchar. optimr updated.
Also displays solution being C(onstrained) or U(nconstrained)
opm() NOT yet updated.
- kktchk: if pHes not symmetric, symmetrize after warning.
- created snewtonmb to handle bounds in a Newton-like method.
This works for a simple function in inst/doc/examples/snewtonmbtest.R
where both bounds and masks appear to work for the example.
However, one correct example is NOT a guarantee of all cases.
snewtonm should be considered EXPERIMENTAL and DIDACTIC for the present.
The author/maintainer would like to be informed of uses, successful or
not, so the code can be improved.
- nmkb is dropped from set of methods for opm() if start is on a bound.
There is a test of this in inst/doc/examples/dropnmbk.R
optimx() does NOT check for this possibility.
- bmchk reviewed
- optimx-package.Rd revised
- Added hessian-uses.R example of Hessian use.
- Add warning if user tries to use optim::SANN in optimx() or optimr()
- demo directory removed. Some checks/tests that are to verify code (but not
needed after code is working) are in inst/doc/examples/
- added testbmstep.R in inst/doc/examples/ to check bmstep(). Minor addition of
comments to bmstep.R
- added starttests to ctrldefault for use in optimx(). There is an example of
use as trystarttests.R in inst/doc/examples/.
- opm() now uses unique(method) to avoid duplicates which crash function.
- Note Rvmminb was missing some bounds resetting steps.
Also some cleanup of details for algorithm controls to use ctrldefault()
e.g., replace dblmax with control$bigval
- grchk and hesschk have ffn1 and ggr1 set to avoid dotargs to avoid name
clash
- axsearch() now gets controls from ctrldefault().
NOTE: THE CALL SYNTAX HAS CHANGED SLIGHTLY.
inst/doc/replace2021/axsearch2018.R has the old version.
- maxfn -- kkt test fixed hjn hjkb. Call to hjn from optimr
uses mcontrol, not control (maximize issue).
- Some cleanup of minor issues of output messages
- Rvmmin, Rvmminu, Rvmminb set to same codes as in Rvmmin "package" updated
December 6, 2021. However in optimr we do NOT use control$maximize.
Maximization is done via optimr().
NOTE: there was an algorithmic error in the earlier codes. FIXED.
- control$checkgrad in Rvmmin defaults to FALSE (was TRUE). In Rcgmin already so.
- Minor improvements to opm.Rd and optimr.Rd regarding control list.
- 'convcode' replaced by 'convergence' in opm.Rd
- lbfgs does NOT report fevals and gevals -- put in optimr documentation
- opm() adjusted so snewton and snewtonm removed from method list when hess()
not provided (==NULL)
- fixed three cases where setting control$maximize=TRUE was failing to set
final function value correctly
- removed printout of trace from Rvmminb()
- minor edits for grammar in snewton.Rd
- moved some tests to ./inst/doc/examples and tidied the files
- optimx routines did not check for NULL gradient or hessian routine gr
and would fail in some instances. Check for missing routines now included,
and optimx.setup forces ugr and uhess NULL if necessary.
- trig1507.R moved from tests to inst/doc/examples/ as the multiple minima
generated are unsuitable as tests
optimx 2020-01-18
- This version does not call some non-CRAN packages.
optimx 2019-12-02
- Replaced all instances of tests on class() with inherits()
- Replaced calls to lbfgsb3 package with those to lbfgsb3c,
and removed calls to the lbfgsb3() function
- Needed to set fval to NA in fnchk when length > 1
optimx 2018-07-10
- Added proptimr() compact display of result of optimr
- Some minor changes to vignettes and to message (msg) output for
line ending
optimx 2018-07-06
- Changed setup of control vector using ctrldefault to allow for use of
individual method defaults
optimx 2018-04-17
- Comments in tests etc. to explain why methods fail
Add hessian to simple test
- Split optim() call in optimx() based on have.bounds. Note glitch in
optim() for vector vs scalar infinite bounds.
- Fix Rvmmin for n=1 infinite loop
optimx 2018-03-30
- Cleanup of output from various solvers and fix some minor omissions
and errors in the returned information
- Conditional output of nlm diagnostic info.
- Changed Rvmmin call. Rvmminb and Rvmminu NOT now called separately,
as user may inadvertently supply inadmissible parameters and get wrong
answer.
- Changed optimr() so nlminb() can use ehess() function
- Replace convcode in optimr() with convergence to align with optim() result
- axsearch() example added to bdstest.R in tests.
- Merged optextras package into optimx
- rationalized optimz-package.R, optimx-package.R into one set of code
- Tests for grchk, kktchk, fnchk included
- Fixed unconditional printout of maxit
- savehess saves (approximate!) hessian with nlminb, call with 'hess' not 'hessian'
optimx 2018-03-28
- Try ! is.numeric(g) vs is.null(g)
- trace changed to integer
optimx 2018-03-18
- added
mcontrol$fnscale <- control$fnscale
in optim() section of optimr for correction of maximization
Thanks to Carlo Lapid [email protected]
- Thanks to playing with examples from Hans Werner Borchers for Nelder-Mead
style codes, found that optimr() did not transfer control$maxit to the
optim() functions. Fixed.
optimx 2018-03-03
- update snewton vignette and tests
- sorted out optimr nlminb failure (used hessian vs hess in optimr call)
optimx 2017-12-15
- merging optimr(x) into optimx
- some documentation of differences between opm() and optimx()
- Use pkg::fn style calls in optimx.run.R
- add bib references to vignette
- checked method="ALL" and control(all.methods=TRUE) are equivalent for opm()
- set trace=0 in all tests
- partially simplify tests
optimx 2017-11-25
- Changed kktchk argument maxfn to maximize to agree with
optimr and other packages
- Tidy tests
- Some cleanup of vignette
optimx 2017-10-05
- Ravi Varadhan as maintainer (reverted later)
- Warning about combining options.
optimx 2017-09-24
- comments include 2017-09-18 version
- brought optimx solvers into optimr and added check that packages
are available
- Fix for lbfgsb3 which was changed at same time to correct and
simplify some control parameters
optimx 2017-7-18
- Simplify example
- Added tests
optimx 2017-04-04
- allowed hessian in opm() and optimr() (for nlm)
- Thanks to Marie Boehnstedt for catching error in creation of
the composite function and gradient for calling nlm() method.
optimx 2017-02-15
optimx 2016-11-30
- Vignette added (incomplete)
optimx 2016-09-22
- account for null gradient
optimx 2016-09-16
- Harry Joe reported troubles with numerical gradient approximations. Found
some glitches and conflicts to fix. Also documentation is less than ideal.
Scaling spar to par was INSIDE control$trace
optimx 2016-08-26
- Thanks to Henrik Singmann for spotting a bug in optimx.run.R
for the case where nlminb fails. I had missed resetting some
elements of the solution object in this case.
optimx 2016-07-10
- hessian=TRUE is NOT properly executed
- Test function hobbs15b illustrates different scalings across all methods
- Appropriate lists of "ALL", "bounded" and "masked" methods are defined
in the ctrldefault() function.
- jonesrun L-BFGS-B fails with numerical gradient approximation.
- dfoptim hjkb was fixed by Ravi Varadhan
optimx 2016-07-03
- Move allmeth, bdmeth, maskmeth into ctrldefault so we have just in
one place. Note that this will change with optimr/optimrx
- Masks for hjn(), and incorporate hjn in optimr
- time (xtimes) removed from optimr, but timing is included in opm()
- polyalgorithm wrapper polyopt() included and documented with example
- multistart wrapper multistart() included and documented with example
- renamed coef.optimx to coef.opm
VERSION optextras 2016-06-30
- Streamlined the package to get rid of ufn, ugr, uhess,
ugHgenb that are so far unused.
- The optsp environment is only currently used
with the simple numerical gradient approximations.
- kktc renamed to kktchk. scalecheck renamed to scalechk
for consistency with other naming.
optimx 2016-6-30
- Streamlined the package to get rid of ufn, ugr, uhess,
ugHgenb that are so far unused.
- The optsp environment is only currently used
with the simple numerical gradient approximations.
- kktc renamed to kktchk. scalecheck renamed to scalechk
for consistency with other naming.
optimx 2016-06-28
- Removed development files from inst, and inst directory
- Checked that controls do not conflict, e.g., maximize and fnscale. Note that
maximize is NOT in optim() call.
- Checked and fixed scaling of bounds in optimr. However, hjkb seems to fail
- Creates scbdtest.R, a test function that is simple but bounded and scaled badly
- Get routines from optextras rather than have duplicates in optimr
- rename package to optimr for optimR(eplacement)
- Problems with lbfgs package (unconstrained) resolved
- test parameter scaling (but this could be improved)
- Demos still refer to optimx -- must be changed -- 160628 moved
off to Roptimtest/current/
- Decided to postpone adding nloptr to set of solvers
optimx 2016-05-30
- Crude setting of uobyqa and newuoa for control rhobeg.
Especially when parameters all 0, or bounds undefined, program
will fail so we need to set this. May also be needed for rhoend.
optimx 2016-01-07
- realized control rhobeg needed to be set or bobyqa would fail.
optimx 2015-11-30
- Changed control "stopbadupdate" to TRUE from FALSE.
optimx 2015-06-03
- Had "lower" instead of "upper" in one of the bounds checks.
Fixed.
- Added index to bstate in optimx.check.R to avoid unwanted
elements in reporting bounds when trace > 0
optimx 2015-05-03
- Check gradient count ig > 2 when doing convergence test.
Possible failure for test fn = sum(x*x) noted. However,
other fns fail with this change. This is an issue in the
program logic rather than an algorithmic failure.
optimx 2015-04-23
- removed condition that made solution NA if gradient or function
limits exceeded. Now returns parameters and function at the point
where limit reached.
optimx 2015-01-21
- Added lbfgsb3, the 2011 version of this code.
optimx 2015-01-07
- comments in codes Rvmminb.R and Rvmminu.R aligned so that the
differences between bounds constrained version and unconstrained
more easily visualized
- control keepinputpar explained. It is used when user does NOT want
out-of-bounds parameters on input moved to nearest bound. Program
will stop in this case.
- Minor changes to maxit and maxfeval defaults so that they are
increased with increasing numbers of parameters.
optimx 2014-12-21
- As per CRAN policies, edited DESCRIPTION and NAMESPACE files.
Because there is no change to code, the date of the package
advanced by only 1 day.
optimx 2014-7-17
- Changed default of control$checkgrad to FALSE. Changed error message
that mentions Rvmmin to Rcgmin.
optimx 2013-11-11
optimx 2013-10-27
- Inserted missing index in bmchk (bdmsk)
- Added maxfevals to control. (It had been there some time.)
optimx 2013-9-24
- Warning that parameter moved to bound suppressed when start
is on the bound.
optimx 2013-09-12
- Modlnp fix from SGN implemented so jones function works.
optimx 2013-9-11
- Fixed some elements of manual. Alias optextras-package
to optextras.
- Fixed some elements of manual. Alias optextras-package
to optextras.
optimx 2013-09-03
- Some fixes, particularly to dot-arguments missing
from calls to user function.
- Changed returned user fgfun to f with g as attribute.
optimx 2013-8-20
- optsp scratchpad environment added. Note that it
is aliased in grfwd to avoid a warning in R CMD
check.
- Can change this in numerical derivative routines
in case of parallel running.
- optsp scratchpad environment added. Note that it
is aliased in grfwd to avoid a warning in R CMD
check.
- Can change this in numerical derivative routines
in case of parallel running.
- Allow quoted gradient names to bring in different gradient
routines. Also to allow gradient approximation step to be
adjusted.
- Dependency on optextras rather than numDeriv
optimx 2013-8-19
- Numerical gradient from optextras, remove bmchk from package,
put bmchk call in Rvmmin, not Rvmminb
optimx 2013-08-08
- Attempt to sort out Rvmmin and Rcgmin no gradient failures.
Resolved by REQUIRING a gradient routine. In optimr() can provide
an approximation via quoted gradient function name.
- Fix for nlm calls which sometimes failed due to inappropriate
control settings and failure to clear (NULL) temporary settings.
optimx 2013-08-07
- Moved optimizer packages to Imports and removed require() calls to
satisfy R 3.1.2.
- This version is artificial! It is the version of 2013-08-06 with an
updated CITATION and minor fixes to satisfy NAMESPACE requirements.
A new optimx is in process with quite sweeping changes that may or
may not result in changed behaviour and usage. (JN 2014-09-04).
- Failure to have the following packages installed may give rise to
diagnostic messages that can be confusing: Rvmmin, Rcgmin, ucminf,
minqa, BB, and dfoptim.
optimx 2013-08-02
- Change startup to avoid warning messages for missing methods.
Add function to check which methods available.
- UOBYQA is present in optimx.run.R, but is not included in the list of
"all.methods" in optimx.setup.R. It should be possible to run UOBYQA
by including "uobyqa" in the list of methods in a direct call to optimx.run.R.
optimx 2013-7-20
- Minor improvements to timing by use of crossprod() in
computing y and D2
optimx 2013-07-09
- patch for missing "details" behaviour
- point to example of use of "follow.on" control
optimx 2013-07-03
- added control usenumDeriv
optimx 2013-06-04
- improved coef() function to extract parameters
optimx 2013-06-03
- Properly packaged version of Rtnmin.
optimx 2013-05-03
- removed optansout.R, get.result.R, get.best.R (no longer needed)
optimx 2013-04-30
optimx 2013-04-29
- removed help pages documenting obsolete features
optimx 2013-04-06
- nmkb and hjkb had no ... arguments in call. Fixed.
- L-BFGS-B (and possibly other optim() methods) return no gevals (count[[2]])
on failure. Forced to NA in case of try-error.
optimx 2013-04-05
- Fixed maximize of function without gradient function. Also glitch
(not tested or encountered) with user supplied Hessian function when
maximizing.
- Separate .Rd files for coef() and summary()
- Use have.bounds to select bounds and unconstrained methods where there
is a choice to avoid duplication of tests. This has been done for
hjk, nmk, Rcgmin, Rvmmin
- Revision of output description in optimx.Rd; some simplifications.
- Parameter names preserved from starting vector, or "p1", "p2", etc. used.
There is a test for these in the ox demo().
optimx 2013-04-03
- in summary.optimx no re-ordering is now specified using order = NULL
- Fixup of examples -- simplified, corrected call to details
optimx 2013-04-02
- in summary.optimx the order= argument may now be an unevaluated
expression (as before), a string representation of that expression or
a character vector of names, e.g. these all work:
order = value # unevaluated expression
order = "value" # character string containing column name
order = list(fevals, gevals) # unevaluated list
order = c("fevals", "gevals") # character vector of column names
order = list(round(value,3), fevals) # unevaluated list
order = "list(round(value,3),fevals)" # expression as string
optimx 2013-03-29
- Note that asking for a method that is NOT in the results will return
a row with rowname (the method name) as "NA". Note that this will NOT
test TRUE with is.na() on the rowname. Use instead a test like
(row.names(ans8missmeth["Rvmmin",])[[1]] == "NA")
- kkt1 and kkt2 in results now returned TRUE/FALSE as in 2012 versions
- Fixed error in kkt2 test (using old name evnhatend rather than hev)
- Results are no longer sorted by optimx, but can be using the 'summary()'
function. This can be called in a variety of ways.
- The 'details' are always kept in a matrix form, even when only one
method has been used.
- 'convcode' is used in place of 'convergence' in result structure
- There is a directory inst/interactive-test that contains ox.R test
script that pauses so user can see interim results.
- As a result of several issues, the result structure is changed from
that of optimx 2012 versions. It is now a data frame with a ‘details’
attribute. Also an 'npar' attribute to give the number of parameters,
and a 'maximize' attribute that is TRUE when the function is to be
maximized.
optimx 2013-03-28
- print.optimx dropped. Not needed as print defaults to print.data.frame.
- added summary.optimx which has order= and par.select= arguments
- order= is an expression or list of expressions using the column names
of the "optimx" object as variables (plus the variable rownames).
- par.select = FALSE now has same meaning as par.select = 0 (rather than
being an error)
- [.optimx now subsets details attribute too
- method column from details attribute dropped and moved to row names
optimx 2013-03-27
- created unit test framework with 2 unit tests in
inst/unitTests/runit.all.R.
To run: demo("unitTests")
optimx 2013-03-25
- [.optimx and as.data.frame.optimx added
- coef.optimx fixed to reflect new ans.ret
optimx 2013-03-22
- ans.ret structure changed
optimx 2013-03-21
- maximize works with print(x, best.only=TRUE)
- KKT checks can be switched off. Decided to NOT separate the code from
optimx.run UNTESTED
- optimx.setup, optimx.run, optimx.check are NOT exported in NAMESPACE,
but a knowledgeable user could use these codes directly, either from
the source code or by modifying the package locally.
optimx 2013-03-18
- removed method= and columns= arguments from print.optimx and added
best.only
- removed print.optimx help page and moved it to the optimx help page
- the row names of the optimx output now show the method and the method
column has been removed
- Removed all user accessible functions except optimx and added
print.optimx.
optimx 2013-03-16
- get.result now emits a warning if method requested is not in
optimx solution
- attempt 2013-3-17 to include summary() in methods to replace
trimpars()
optimx 2013-2-20
- Removed fail, setstep (not used), ceps from Rcgminu (not used)
optimx 2012-8-08
- Minor edits and addition of references.
optimx 2012-07-18
- Extracted helper routines from optplus to
create optextras package to put on CRAN.
Formerly in optplus or optfntools on R-forge,
but difficulties with dependencies indicated
that the support routines need to be on CRAN.
Renamed to avoid name confusion.
optimx 2012-7-18
- Extracted helper routines from optplus to
create optextras package to put on CRAN.
Formerly in optplus or optfntools on R-forge,
but difficulties with dependencies indicated
that the support routines need to be on CRAN.
Renamed to avoid name confusion.
optimx 2012-07-09
- Suppress warnings in ugHgenb -- too much
unhelpful output
- Fixed optansout to print parameters
optimx 2012-7-09
- Suppress warnings in ugHgenb -- too much
unhelpful output
- Fixed optansout to print parameters
optimx 2012-7-9
- Removed optansout to avoid conflicts with optplus/optimx
- Changes before 2012-4-30 in Rvmmin not recorded
optimx 2012-6-09
- User-tolerance for gradient size.
optimx 2012-06-07
- Merged optfntools into optplus to ensure
R-forge build.
optimx 2012-6-07
- Merged optfntools into optplus to ensure
R-forge build.
optimx 2012-05-26
- Added check for "on a bound" in bmchk to aid
in use of tanh() transformation as in nmkb().
- scalechk takes account of masks (masked parameters not used for
computing scale)
- Changes before 2012-5-26 in the optextras package not recorded
optimx 2012-5-26
- Added check for "on a bound" in bmchk to aid
in use of tanh() transformation as in nmkb().
optimx 2012-05-21
- Separation of constrained and unconstrained versions.
optimx 2012-5-21
- Separation of constrained and unconstrained versions.
- Changes in Rcgmin before 2012-4-30 not recorded
TODOS in Rcgmin or ncg or Rvmmin or nvm
- Eventually include different line searches, possibly in a
separate routine