lines 6-761 of file: appendix/whats_new/19.xrst

{xrst_begin whats_new_19 app}
{xrst_spell
   autotools
   bthread
   cholesky
   conditionally
   deprecate
   entires
   lepp
   mmdd
   omhelp
   simplification
   subgraphs
   uninitialized
   usr
   yy
}

Changes and Additions to CppAD During 2019
##########################################

mm-dd
*****

12-27
=====

#. Add the :ref:`graph_op_enum@Print`
   operator to the :ref:`cpp_ad_graph-name` and :ref:`json_ad_graph-name` ; see
   :ref:`graph_print_op.cpp-name` , :ref:`json_print_op.cpp-name` .
#. Improve the error messaging when a Json
   :ref:`json_ad_graph@op_define_vec@op_code` in the
   operator definition section does not start with one
   and increment by one between definitions.

12-25
=====

#. Change *pos* to :ref:`PrintFor@notpos`
   in the ``PrintFor`` documentation because the text is printed
   when it is not positive.
#. Add C++ graph member functions that :ref:`cpp_graph_vector@Syntax@Find`
   the index for an atomic or discrete function.

12-24
=====

#. Add the :ref:`discrete function<json_graph_op@Discrete Functions>`
   operator to the :ref:`json_ad_graph-name` .
#. Fix bug when using a dynamic parameter for
   :ref:`PrintFor@notpos` or :ref:`PrintFor@value`
   in the ``PrintFor`` routine.

12-23
=====
Add the :ref:`graph_op_enum@Discrete Function`
operator to the :ref:`cpp_ad_graph-name` ; see
:ref:`graph_discrete_op.cpp-name` .

12-21
=====
Add an
:ref:`wish_list@Abs-normal@Atomic Functions` item
to the wish list for abs-normal functions.

12-20
=====
Add a
:ref:`wish_list@Dynamic Parameters` item and a
:ref:`wish_list@Graph Operators` item to the wish list.

12-18
=====

#. The :ref:`graph_atom_op.cpp-name` *Base* type was changed from
   ``double`` to ``float`` .
#. Fix a bug in
   :ref:`from_graph-name` , :ref:`from_json-name` , :ref:`to_graph-name` , and :ref:`to_json-name`
   when the :ref:`from_json@Base` type is not ``double`` .

12-16
=====
Add more description in the comments for the
:ref:`graph_op_enum values<graph_op_enum@Enum Values>` .

12-15
=====

#. Added a list of the possible :ref:`ADFun-name` operators
   that are :ref:`missing<graph_op_enum@Missing Operators>`
   (not yet implemented) by :ref:`to_graph-name` and :ref:`to_json-name` .
#. Absolute zero multiplication
   :ref:`json_graph_op@Binary Operators@azmul`
   was as added to the :ref:`json_ad_graph-name` and
   :ref:`cpp_ad_graph-name` as ``pow_graph_op`` .
   Functions that contain the this operator
   can be converted to and from ``cpp_graph`` objects; e.g.,
   see :ref:`graph_azmul_op.cpp-name` .
   They can also be converted to and from strings containing
   :ref:`json_ad_graph-name` representations; e.g.,
   see :ref:`json_azmul_op.cpp-name` .
#. Improve the
   :ref:`graph_op_enum examples<graph_op_enum@Examples>` by removing
   some type definitions and using statements.
   In addition, use :ref:`cppad_testvector-name` to demonstrate that
   any :ref:`simplevector-name` type can be used.

12-14
=====
The power operators was added to the :ref:`json_ad_graph-name` as
:ref:`json_graph_op@Binary Operators@pow` and to the
:ref:`cpp_ad_graph-name` as ``pow_graph_op`` ; see
:ref:`graph_op_enum@Enum Values` .
Functions that contain the :ref:`pow-name` operators
can be converted to and from ``cpp_graph`` objects; e.g.,
see :ref:`graph_pow_op.cpp-name` .
They can also be converted to and from strings containing
:ref:`json_ad_graph-name` representations; e.g.,
see :ref:`json_pow_op.cpp-name` .

12-13
=====
Add :ref:`switch_var_dyn.cpp-name` , an example that
switches between variables and dynamic parameters.

12-12
=====
Add a version of the ``from_graph`` routine that
can convert dynamic parameters to variables and vice versa; see
:ref:`from_graph@dyn2var` and *var2dyn* in the
``from_graph`` documentation.

12-09
=====
Add the ``Independent`` ( *x* , *dynamic* ) syntax
to the :ref:`Independent-name` function and changes some of the
:ref:`new_dynamic-name` example to use this syntax.

12-08
=====
Add wish list item for converting dynamic parameters to variables
and vice versa. This has since been completed;
see :ref:`whats_new_19@mm-dd@12-12` above.

11-15
=====
Fix clang 9.0.0 mistaken uninitialized warning in
:ref:`a11c_openmp.cpp-name` .

11-10
=====
Change *n_independent* to *n_variable_ind* in
:ref:`json_ad_graph<json_ad_graph@variable_ind_vec@n_variable_ind>` and
:ref:`cpp_ad_graph<cpp_ad_graph@n_variable_ind>` .
In addition, the under construction warning was removed from
:ref:`json_ad_graph-name` and :ref:`cpp_ad_graph-name` .

11-09
=====
The C++ data structure for an AD graph :ref:`cpp_ad_graph@cpp_graph`
was add the user API for CppAD.

11-05
=====
More work on reducing the amount of memory used by the
:ref:`from_json-name` and :ref:`to_json-name` operations.

10-28
=====
More work on reducing the amount of memory used by the
:ref:`from_json-name` and :ref:`to_json-name` operations.

10-27
=====

#. Working on reducing the amount of memory used by the
   :ref:`from_json-name` and :ref:`to_json-name` operations.
#. Improve sparse jacobian error messaging when coloring is
   :ref:`sparse_jac@coloring@cppad` and
   :ref:`sparse_jac@subset`
   has elements that are not in the sparsity
   :ref:`sparse_jac@pattern` .

10-22
=====
A left (right) bracket was added to the beginning of (end of) the
:ref:`json_ad_graph@constant_vec` ,
:ref:`json_ad_graph@op_usage_vec` ,
:ref:`json_ad_graph@dependent_vec` values
in a Json AD graph (to properly conform to Json format).

10-21
=====
Move the Json AD graph discussion of
:ref:`json_ad_graph@Node Indices`
under one heading and include
a diagram of all the nodes.

10-17
=====
Include conversion from function to graph (:ref:`to_json-name` )
and back to function (:ref:`from_json-name` )
in the :ref:`json operator<json_graph_op-name>` examples.

10-02
=====
The default value for :ref:`cmake@cppad_prefix`
was changed from ``/usr`` to the value of the cmake variable
`CMAKE_INSTALL_PREFIX <https://cmake.org/cmake/help/v3.0/variable/CMAKE_INSTALL_PREFIX.html>`_.

09-30
=====
Use a newer version of the program that generates the documentation
``omhelp`` .
The search utility now shows the title and section tag in separate
areas and hides the other keywords.
Hitting return or double clicking in the title area,
in addition to choosing the ``Goto`` button, will go to
the selected section.

09-24
=====
Discussion of a :ref:`cmake@CMake Command@Warning`
about boost as added to cmake command documentation.

09-23
=====
The ``cppad_deprecated`` option has been completely removed from the
cmake command :ref:`cmake@CMake Command@Options` documentation.

09-18
=====
Merge in changes from json branch which implement the
:ref:`json_graph_op@Compare Operators` .

09-17
=====
The autotools install was installing * . ``hpp.in``
instead of the corresponding * . ``hpp`` file.
This has been fixed.

09-14
=====
Fix autotools install so files get installed to
*prefix* / ``include/cppad`` instead of
*prefix* / ``include/include/cppad`` .
Also fix the autotools ``make test`` .

09-12
=====
Bring :ref:`configure-name` build and test up to date
(this was deprecated at end of 2012 and support for this will eventually stop).

09-10
=====
The old Json unary operator examples were converted to be just
tests and a single :ref:`json_unary_op.cpp-name` example was left
in its place.

09-09
=====
The Json graph
:ref:`json_graph_op@Conditional Expressions`
were added.

08-29
=====

#. Extend :ref:`to_json-name` to handle
   :ref:`json_graph_op@Atomic Functions` .
   This should have been included with the atomic function change
   on :ref:`whats_new_19@mm-dd@08-17` .
#. Simplify the :ref:`json_atom_op.cpp-name` and add a call
   to :ref:`to_json-name` to this example.
#. The ``string_vec`` field was removed from the definition of an
   Json AD graph :ref:`json_ad_graph@AD Graph@function` .
   The corresponding operators will have strings in their
   :ref:`op_usage<json_ad_graph@op_usage@Strings In Usage>` .

#. The vector branch was merged into the master.
   This brought in the changes below

#. 08-25:
   Change :ref:`cppad_vector.cpp-name` example to use same names as
   in :ref:`CppAD_vector-name` documentation.

#. 08-23:
   The :ref:`CppAD_vector-name` documentation page was edited (improved).

#. 08-22:
   :ref:`CppAD_vector@Iterators` were added
   to the CppAD vector class.

08-26
=====
The Json graph
:ref:`json_graph_op@Unary Operators` were added.

08-19
=====
The CppAD vector
:ref:`CppAD_vector@Include` discussion was edited.
In addition, the class :ref:`CppAD_vector@vectorBool`
is now available as a separate include file.

08-17
=====

#. Add a the Json addition example :ref:`json_add_op.cpp-name` .
#. Add the :ref:`atomic function<json_graph_op@Atomic Functions>`
   operator which enables
   one to call atomic functions as part of a Json AD graph; see the
   :ref:`json_atom_op.cpp-name` example.

08-16
=====
Add the case where :ref:`atomic_three_forward@need_y` is equal to
``constant_enum`` during ``atomic_three`` forward mode.
This only needs to be implemented when by :ref:`from_json-name` is used
with the atomic function.

08-13
=====
Fix a bug in the calculation of
Hessian :ref:`sparsity patterns<sparsity_pattern-name>`
when using :ref:`chkpoint_two-name` function.

08-12
=====

#. Fix an assert error for the
   :ref:`json_graph_op@sum` operator
   in :ref:`from_json-name` .
#. Improve the documentation for mixing debug and release versions
   of CppAD in the same program using ``CPPAD_DEBUG_AND_RELEASE`` .

08-10
=====

#. Remove under construction heading from the examples
   :ref:`json_mul_op.cpp-name` , :ref:`json_sub_op.cpp-name` , :ref:`json_sum_op.cpp-name`
   (they have been working for a while now).
#. Add the Json :ref:`json_graph_op@Binary Operators@div` operator.

08-09
=====

#. Add :ref:`json_ad_graph@AD Graph@function_name`
   to the Json graph representation of a function.
#. Use the function name for reporting errors in a Json graph.

08-08
=====

#. Move the example include files
   ``eigen_cholesky.hpp`` ,
   ``eigen_mat_inv.hpp`` ,
   ``eigen_mat_mul.hpp``
   from the ``include/cppad/example`` directory
   to the ``include/cppad/example/atomic_two`` directory.
#. Move the example include file
   ``mat_mul.hpp``
   from the ``include/cppad/example`` directory
   to the ``include/cppad/example/atomic_three`` directory.
#. Improve the :ref:`atomic_three_define@Syntax` ,
   and :ref:`atomic_three_define@parameter_x` discussion,
   in the documentation for the ``atomic_three`` functions.
   Furthermore, add an extra discussion of the parameters
   to the documentation for *taylor_x* in the
   :ref:`forward<atomic_three_forward@taylor_x@parameters>` and
   :ref:`reverse<atomic_three_forward@taylor_x@parameters>` callback functions.
#. Change Json *string_vec*
   from first to second line below:

   | |tab| [ *n_string* , [ *first_string* , ..., *last_string*  ] ]
   | |tab| *n_string* , [ *first_string* , ..., *last_string*  ]

   ( *string_vec* was removed on :ref:`whats_new_19@mm-dd@08-29` ).
#. Change Json :ref:`json_ad_graph@constant_vec`
   from first to second line below:

   | |tab| [ *n_constant* , [ *first_constant* , ..., *last_constant*  ] ]
   | |tab| *n_constant* , [ *first_constant* , ..., *last_constant*  ]

#. Change Json :ref:`json_ad_graph@op_usage_vec`
   from first to second line below:

   | |tab| [ *n_op_usage* , [ *first_op_usage* , ..., *last_op_usage*  ] ]
   | |tab| *n_op_usage* , [ *first_op_usage* , ..., *last_op_usage*  ]

#. Change Json :ref:`json_ad_graph@dependent_vec`
   from first to second line below:

   | |tab| [ *n_dependent* , [ *first_dependent* , ..., *last_dependent*  ] ]
   | |tab| *n_dependent* , [ *first_dependent* , ..., *last_dependent*  ]

08-06
=====

#. Fix a bug in ``chkpoint_two`` when
   :ref:`chkpoint_two_ctor@use_in_parallel` is true.
#. Add the :ref:`multi_atomic_three.cpp-name` and :ref:`multi_chkpoint_two.cpp-name`
   examples.
#. Deprecate the :ref:`multi_atomic_two.cpp-name` and :ref:`multi_chkpoint_one.cpp-name`
   examples

07-31
=====
Add an example representation of a sparse matrix using the an
:ref:`json_ad_graph-name` ; see :ref:`json_sparse.cpp-name` .

07-30
=====

#. Add the json subtraction operator
   :ref:`json_graph_op@Binary Operators@sub` .
#. Fix :ref:`from_json-name` addition and multiply operations
   where the first argument is a variable and second argument is a parameter.
#. Add simple examples for the json operators that have been implemented; e.g.,
   :ref:`json_sum_op.cpp-name` .
#. Fix :ref:`to_json-name` when a cumulative summation has
   subtraction terms in its summation.

07-29
=====
Add the :ref:`json_sum_op.cpp-name` example and fix a bug in
:ref:`from_json-name` that was demonstrated by this example.
To be specific, the json :ref:`json_graph_op@sum` operator
did not work when all its result was a
:ref:`dynamic parameter<glossary@Parameter@Dynamic>` .

07-25
=====
Fix the warning (on some systems) that
:ref:`rev_jac_sparsity-name` shadows a member of
``CppAD::atomic_three<Base>`` .

07-19
=====
The ``cppad_lib`` dynamic link library version number used to be
*yy* . *mmdd* where the major version number *yy*
was the year minus 2000 and the minor version number *mmdd*
was the month and day.
The value *mmdd* was to large for a Mac system minor version number.
See :ref:`cmake@cmake_install_libdirs@cppad_lib` for the new
library version number system.

07-18
=====
Change some assignment operators to equality operators in a few asserts
(``=`` to ``==`` in the files
``optimize_run.hpp`` and ``player.hpp`` ).
These were not bugs because the corresponding variable
was being set to the value it should have been.

07-05
=====

#. The :ref:`json_ad_graph@op_usage_vec@n_usage` value generated by
   :ref:`to_json-name` was not correct. This has been fixed.
#. All the operators that were implemented so far were included in the
   :ref:`json_ad_graph@op_define_vec` section generated by
   ``to_json`` .
   Now only operators that are used are included.
   This makes it easier for humans to understand simple example graphs
   (especially as the number of implemented operators increases).
#. There was a bug in :ref:`from_json-name` when one
   of the operations was a dynamic parameter and the other was a constant.
   This has been fixed.
#. All of the addition and multiplication operators
   (including cumulative sum operators) have been implemented and tested..

07-04
=====

#. Improve the :ref:`from_json-name` error messages by including the
   previous two lines of Json in error message.
#. Working on a :ref:`to_json-name` example that outputs graph
   for the derivative of a function.

07-03
=====
The format of a Json computational graph was changed to include
a section at the top that defines the mapping
between integer codes in the graph and corresponding operators,
see the heading :ref:`json_ad_graph@op_define_vec` .
This is now separate from
:ref:`json_ad_graph@op_usage_vec` which is the part
of the graph that is expect to be very large and is now less verbose.

06-29
=====
It used to be the case that optimized functions with cumulative
summation operators could not be optimized.
This has been fixed and it is now acceptable to
:ref:`optimize@Re-Optimize` an ``ADFun`` object.

06-19
=====

#. Some work was done to improve the speed of sparsity calculations
   when using the internal representation for a
   :ref:`glossary@Sparsity Pattern@Boolean Vector` .
#. An extra pair of braces was expected in each
   :ref:`json_ad_graph@op_usage` of an :ref:`json_ad_graph-name` .
   This has been fixed.

06-18
=====
There are some problems using as Windows DLL for the ``cppad_lib``
library. These problems have been circumvented (for the time being) by using
as static library for Windows systems.

06-17
=====
A new feature that constructs a ``ADFun`` object
from a Json representation of a computational graph was added; see
:ref:`json_ad_graph-name` and :ref:`from_json-name` .
This requires linking the ``cppad_lib`` library.
It is a Minimal Viable Product and is intended for testing.
As of yet, only has the addition and multiplication operators are
implemented.

06-10
=====
Some work was done to improve the speed of sparsity calculations
when using the internal representation for a
:ref:`glossary@Sparsity Pattern@Vector of Sets` .
This also reduced the time required by the
:ref:`optimize-name` operation for some cases.

06-02
=====
The version number in the title of the top level for the
:ref:`user_guide-name` documentation was not being updated.
This started on ``20190521`` and has been fixed.

05-31
=====
Add the :ref:`swap<CppAD_vector-name>` operation to the
``CppAD::vector`` template class and
:ref:`CppAD_vector@vectorBool` class.
In addition, change :ref:`cppad_vector.cpp-name` example
so it focuses on operations that are not in
:ref:`SimpleVector-name` .

05-28
=====
There was a bug in optimizing conditional expressions; i.e.,
optimizing without the option
:ref:`optimize@options@no_conditional_skip` .
This was a rare case that involving sets of conditional expressions
and conditionally skipping a specific operator.
The bug was reproduced in a regression test and is fixed.

05-22
=====

#. Some speed improvements were made to the
   :ref:`for_hes_sparsity-name` algorithm.
   To be specific the forward Jacobian and Hessian patterns
   are computed during the same pass and using the same structure.
   This avoids an extra pass on the operation sequence.
   In addition, when :ref:`for_hes_sparsity@internal_bool`
   is false, it enables smart points to the forward Jacobian and Hessian
   to share the same underlying sets of integers (when they are equal).
#. Some minor improvements and updating was done for the
   :ref:`directory-name` and :ref:`glossary-name` sections.

05-21
=====
Simplify the example computation of Hessian using subgraphs
:ref:`subgraph_hes2jac.cpp-name` .
This simplification was a change from using
:ref:`multiple levels of AD<mul_level-name>` to using :ref:`base2ad-name` .

05-07
=====
There was a bug in the checking for correctness of the
:ref:`adolc_sparse_hessian.cpp-name` results.
This was due to the fact that the sparsity patterns for ADOL-C
may have more entires that CppAD knows are zero.

04-01
=====
Add the optional
:ref:`time_test@repeat_out` argument to the
``time_test`` function.

03-27
=====
Improve the speed of sparse Hessian calculations that
:ref:`sparse_hes@coloring@colpack.symmetric`
coloring method by directly accessing the coloring determined by Colpack.

03-19
=====
There was a bug in :ref:`f.Reverse(q, w)<reverse_any-name>` when *f*
contained the  :ref:`pow(x,y)<pow-name>` function and *x*  == 0 .
To be specific, it was not reporting zero instead of ``nan``
for the derivative. This has been partially, but not completely fixed.
See the :ref:`wish_list@Reverse Mode` in the wish list.

03-13
=====
Fix a bug in :ref:`f.subgraph_jac_rev(x, subset)<subgraph_jac_rev-name>` .
This bug appeared in a test that only returned the upper triangle
of a symmetric matrix (the lower triangle worked).

02-26
=====
There was a problem linking the boost multi-threading library
on some systems that caused the corresponding check to be skipped.
This has been fixed. To be specific,
::

   make check_example_multi_thread_bthread: available

was not in the :ref:`cmake@CMake Command` output.
This has been fixed.

02-23
=====
The return value *n_sweep* was changed to
:ref:`sparse_jac@n_color` because it is not the number
of sweeps when :ref:`sparse_jac@group_max` is not equal to one.
This change was also made for the
:ref:`speed tests<speed_main@Speed Results@n_color>` .

02-22
=====
Add row major order specifications to the speed test routines
:ref:`link_sparse_jacobian<link_sparse_jacobian@Row Major>` and
:ref:`link_sparse_hessian<link_sparse_hessian@Row Major>` .
This reduced the overhead (increased the speed) for the
:ref:`adolc_sparse_jacobian.cpp-name` and
:ref:`adolc_sparse_hessian.cpp-name` test cases.

02-20
=====

#. Fix bug below in the new release
   `20190200.1 <https://github.com/coin-or/CppAD/archive/20190200.1.tar.gz>`_.
#. There was a problem locating the boost multi-threading library.
   To be specific, the
   :ref:`cmake@CMake Command` would report

      ``Could not find boost multi-threading library``

   but it would still try to use the library.
   This was because a file name with a ``-`` in it was
   also being used for a cmake variable name.
   This has been fixed.

02-06
=====
Improve the discussion of the purpose of for the different
:ref:`subgraph<subgraph_jac_rev@Purpose>` options which computing
sparse Jacobians.

02-01
=====

#. A stable release for 2019 is now available on the
   `release <https://github.com/coin-or/CppAD/releases>`_ page.
#. The :ref:`download-name` page was edited to include better discussion
   of the license and documentation.

01-18
=====
The documentation for :ref:`atomic_three-name` ,
and its subsections, was improved.

01-17
=====
The arguments
:ref:`atomic_three_define@parameter_x` and
:ref:`atomic_three_define@type_x` were included
for all the ``atomic_three`` virtual functions.

01-16
=====
More improvements, and bug fixes, in the optimization of :ref:`atomic-name`
operations.
This results in a reduction in the number of parameters
:ref:`fun_property@size_par`
and the number of variables
:ref:`fun_property@size_var` .

01-15
=====
Fix a bug in the optimization of :ref:`atomic_three-name`
operations with :ref:`Independent@dynamic` parameters.

01-14
=====

#. The :ref:`rev_depend<atomic_three_rev_depend-name>` call back function
   was added to ``atomic_three`` class.
   This enables one to :ref:`optimize-name` functions that use
   ``atomic_three`` operations; see
   the :ref:`rev_depend.cpp<atomic_three_rev_depend.cpp-name>` example.
#. The ability to :ref:`optimize-name` functions was extended to
   :ref:`chkpoint_two-name` operations; e.g., a call to optimize was added
   at the end of the
   :ref:`get_started.cpp<chkpoint_two_get_started.cpp-name>` example.
#. The :ref:`chkpoint_one-name` class was deprecated,
   use :ref:`chkpoint_two-name` instead.

01-06
=====
One of the two atomic matrix multiply examples was converted
from using :ref:`atomic_two-name` to :ref:`atomic_three_mat_mul.cpp-name` .
This conversion made it clear that the *type_x* argument
needed to be included in the
:ref:`jac_sparsity<atomic_three_jac_sparsity@type_x>` and
:ref:`hes_sparsity<atomic_three_hes_sparsity@type_x>` functions.

01-05
=====
Merge in the ``atomic branch`` .
This completed :ref:`wish_list-name` item for the new atomic function API;
see :ref:`atomic_three-name` .
This is an improvement over the :ref:`atomic_two-name` interface,
which used the ``atomic_base`` class and has been deprecated.
In addition the following changes were included:

#. Move the checkpoint examples below the
   :ref:`checkpoint<chkpoint_one-name>` documentation.
#. A phantom parameter, at index zero, was added; see
   :ref:`fun_property@size_par` .
#. There appears to have been a bug in ``put_con_par`` (not tested for)
   whereby a constant parameter might match a dynamic parameter
   (and get changed). This has been fixed.
#. There was a mistake in the ``check_variable_dag`` routine whereby
   the operator ``LeppOp`` checked a parameter index as if it were
   a variable index. This has been fixed.

01-11
=====
Merge in the ``chkpoint`` branch.

#. This created the :ref:`chkpoint_two-name` class as a replacement
   for the :ref:`chkpoint_one-name` class.
   This new checkpoint class takes advantage
   of :ref:`dynamic parameters<glossary@Parameter@Dynamic>`
   and can be used with :ref:`base2ad-name` .
#. While creating the ``chkpoint_two`` class,
   a problem was discovered in the :ref:`atomic_three-name` class.
   To be specific, it does not yet work with :ref:`optimize-name` .
   A reverse dependency analysis is the proper way to fix this.
   The ``atomic_three`` ``type`` member function does a forward
   :ref:`atomic_three_for_type@Dependency Analysis`
   and so its name was changed from ``type`` to ``for_type`` .

{xrst_end whats_new_19}