Blurb::
Pattern search, derivative free optimization method
Description::
The asynchronous parallel pattern search (APPS) algorithm
\cite GrKo06 is a fully asynchronous pattern search
technique in that the search along each offset direction continues
without waiting for searches along other directions to finish.

Currently, APPS only supports coordinate bases with a total of \e 2n
function evaluations in the pattern, and these patterns may only
contract.

<b> Concurrency </b>

APPS exploits parallelism through the use of %Dakota's concurrent
function evaluations. The variant of the algorithm that is currently
exposed, however, limits the amount of concurrency that can be
exploited. In particular, APPS can leverage an evaluation concurrency
level of at most twice the number of variables. More options that
allow for greater evaluation concurrency may be exposed in future
releases.

<b> Algorithm Behavior </b>

\li \c initial_delta: the initial step length, must be positive
\li \c variable_tolerance: step length used to determine convergence,
must be greater than or equal to 4.4e-16
\li \c contraction_factor: amount by which step length is rescaled
after unsuccesful iterates, must be strictly between 0 and 1


<b> Merit Functions </b>

APPS solves nonlinearly constrained problems by solving a sequence of
linearly constrained merit function-base subproblems. There are
several exact and smoothed exact penalty functions that can be
specified with the \c merit_function control. The options are as
follows:

\li \c merit_max: based on \f$ \ell_\infty\f$ norm
\li \c merit_max_smooth: based on smoothed \f$ \ell_\infty\f$ norm
\li \c merit1: based on \f$ \ell_1\f$ norm
\li \c merit1_smooth: based on smoothed \f$ \ell_1\f$ norm
\li \c merit2: based on \f$ \ell_2\f$ norm
\li \c merit2_smooth: based on smoothed \f$ \ell_2\f$ norm
\li \c merit2_squared: based on \f$ \ell_2^2\f$ norm

The user can also specify the following to affect the merit functions:
\li \c constraint_penalty
\li \c smoothing_parameter


<b> Method Independent Controls </b>

The only method independent controls that are currently mapped to APPS
are:
\li \ref method-asynch_pattern_search-max_function_evaluations
\li \ref method-asynch_pattern_search-constraint_tolerance
\li \ref method-output

Note that while APPS treats the constraint tolerance separately for
linear and nonlinear constraints, we apply the same value to both if
the user specifies \c constraint_tolerance.

The APPS internal display level is mapped to
the %Dakota \c output settings as follows:
\li \c debug: display final solution, all input parameters,
variable and constraint info, trial points, search directions, and
execution details
\li \c verbose: display final solution, all input parameters,
variable and constraint info, and trial points
\li \c normal: display final solution, all input parameters,
variable and constraint summaries, and new best points
\li \c quiet: display final solution and all input parameters
\li \c silent: display final solution



<b>Expected HDF5 Output</b> 

If Dakota was built with HDF5 support and run with the 
\ref environment-results_output-hdf5 keyword, this method 
writes the following results to HDF5: 

- \ref hdf5_results-best_params 
- \ref hdf5_results-best_obj_fncs (when \ref responses-objective_functions) are specified) 
- \ref hdf5_results-best_constraints 
- \ref hdf5_results-calibration (when \ref responses-calibration_terms are specified) 

Topics::	package_hopspack, global_optimization_methods
Examples::
Theory::
Faq::
See_Also::	
