libcmaes 0.10.2
A C++11 library for stochastic optimization with CMA-ES
|
profile likelihood object holder as a set of points and values. More...
#include <libcmaes/pli.h>
Public Member Functions | |
pli (const int &k, const int &samplesize, const int &dim, const dVec &xm, const double &fvalue, const double &fup, const double &delta) | |
profile likelihood constructor | |
std::pair< double, double > | getMinMax (const double &fvalue, int &minindex, int &maxindex) |
find bounds around the objective function parameters for a given value of f, base on pre-computed profile likelihood points. | |
void | setMinMax () |
void | setErrMinMax () |
sets the errors bounds for this profile likelihood. | |
double | get_err_min () const |
get lower error bound | |
double | get_err_max () const |
get upper error bound | |
int | get_k () const |
int | get_samplesize () const |
dVec | get_fvaluem () const |
dMat | get_xm () const |
double | get_min () const |
double | get_max () const |
Private Attributes | |
int | _k = -1 |
int | _samplesize = 0 |
dVec | _fvaluem |
dMat | _xm |
double | _min = 0.0 |
double | _max = 0.0 |
double | _errmin = 0.0 |
double | _errmax = 0.0 |
int | _minindex = -1 |
int | _maxindex = -1 |
std::vector< int > | _err |
double | _fup |
double | _delta |
Friends | |
class | CMASolutions |
template<class U > | |
class | errstats |
profile likelihood object holder as a set of points and values.
|
inline |
profile likelihood constructor
k | dimension in which the profile likelihood was computed |
samplesize | number of steps of the linesearch direction |
dim | dimension of the objective function parameter space |
xm | vector of parameters at fvalue |
fvalue | the function minima around which the profile likelihood was computed |
fup | the function deviation for which the profile likelihood was computed |
delta | tolerance around fvalue + fup for which the profile likelihood was computed |
|
inline |
get upper error bound
|
inline |
get lower error bound
|
inline |
find bounds around the objective function parameters for a given value of f, base on pre-computed profile likelihood points.
fvalue | function value |
minindex | index of the profile likelihood point that is the lower bound |
maxindex | index of the profile likelihood point that is the upper bound |
|
inline |
\brie sets the bounds for this profile likelihood object based on original function value + fup