Diana  0.8.3
Public Member Functions | Public Attributes
ExperimentSystem Class Reference

#include <DiffitModel.hpp>

List of all members.

Public Member Functions

 ExperimentSystem (const long _nvar, const long _npar, const long _nobs)
 ExperimentSystem (const ExperimentSystem &ex)
void updateResidualsFrom (const ExperimentSystem &_sys)

Public Attributes

int * y0fix
int * doP
long nobs
 Number of observed quantities.
long nvar
 Number of variables/equations.
long npar
 Number of parameters.
long nMeasure
 Number of measured data.
double * xMeasure
 Temporal data points.
double ** sigma
 Standard deviation of data.
double fitstart
 Fitting range.
double fitend
long firstMeasure
 First point in fitting range.
long lastMeasure
 Last point in fitting range.
long nPoints
 Number of multiple shooting intervals.
double * mesh
 Multiple shooting mesh.
long me
 Number of equality constraints.
long mg
 Number of inequality constraints.
double objF
 Current value of the objective function.
double ** h
 Discrepancies/deviation form continuous trajectories.
double ** residues
 Current residues.
double * r2
 Right-hand side of equality constraints.
double * r3
 Right-hand side of inequality constraints.
double *** dyds
 Derivative of the trajectory with respect to intial values inside the multiple shooting interval.
double *** dydp
 Derivative of the trajectory with respect to parameters inside the multiple shooting interval.
double *** dmds
 Derivative of the observed quantities with respect to intial values inside the multiple shooting interval.
double *** dmdp
 Derivative of the observed quantities with respect to parameters inside the multiple shooting interval.
double *** dR2ds
 Derivative of the equality constraints with respect to intial values inside the multiple shooting interval.
double ** dR2dp
 Derivative of the equality constraints with respect to parameters.
double *** dR3ds
 Derivative of the inequality constraints with respect to intial values inside the multiple shooting interval.
double ** dR3dp
 Derivative of the inequality constraints with respect to parameters.
double * ua
 Condensation of the least squares.
double ** Ea
double ** Pa
double * ue
 Condensation of the equality constraints.
double ** Ee
double ** Pe
double * ug
 Condensation of the inequality constraints.
double ** Eg
double ** Pg
double * par
 Parameters.
double ** dS
double * dP
double * y0
 Initial value.
double ** yComp
 Predicted trajectory (intergation)
double ** yPred
 Predicted observables (intergation)
double ** yTry
 Initial guesses at mesh points.
double ** yMeasure
 Data of the observations.
char * fileName
 File name of experiment.
string * splineFile
double ** splineNodes
 Spline nodes.
double ** splineY
 Value at the spline nodes.
double ** splineGam
 Second derivative at the spline nodes.
unsigned long * nNodes
 Number of the spline nodes.
int splinesDefined
 Flag: TRUE if splines are defined.
long NSPLINES

Detailed Description

ExperimentSystem defines the data for a linear equationset, based on comparison of experiments with model predictions


Member Data Documentation

double*** dmdp

Derivative of the observed quantities with respect to parameters inside the multiple shooting interval.

3-tensor of dimension:

\[ {\bf nMeasure} \times {\bf nobs} \times {\bf npar} \]

double*** dmds

Derivative of the observed quantities with respect to intial values inside the multiple shooting interval.

3-tensor of dimension:

\[ {\bf nMeasure} \times {\bf nobs} \times {\bf nvar} \]

double* dP

Parameter update

double** dR2dp

Derivative of the equality constraints with respect to parameters.

Matrix of dimension:

\[ {\bf me} \times {\bf npar} \]

double*** dR2ds

Derivative of the equality constraints with respect to intial values inside the multiple shooting interval.

3-tensor of dimension:

\[ {\bf me} \times {\bf nPoints} \times {\bf nvar} \]

double** dR3dp

Derivative of the inequality constraints with respect to parameters.

Matrix of dimension:

\[ {\bf mg} \times {\bf npar} \]

double*** dR3ds

Derivative of the inequality constraints with respect to intial values inside the multiple shooting interval.

3-tensor of dimension:

\[ {\bf mg} \times {\bf nPoints} \times {\bf nvar} \]

double** dS

Initial states update

double*** dydp

Derivative of the trajectory with respect to parameters inside the multiple shooting interval.

3-tensor of dimension:

\[ {\bf nPoints} \times {\bf nvar} \times {\bf npar} \]

double*** dyds

Derivative of the trajectory with respect to intial values inside the multiple shooting interval.

3-tensor of dimension:

\[ {\bf nPoints} \times {\bf nvar} \times {\bf nvar} \]

double** h

Discrepancies/deviation form continuous trajectories.

\[{\bf h}_{i,j}={\bf yComp}_{i,j}-{\bf yTry}_{i,j} \]

Matrix of dimension:

\[ {\bf nPoints}\times{\bf nvar}\]

double* mesh

Multiple shooting mesh.

Vector of dimension:

\[ {\bf nPoints} \]

long npar

Number of parameters.

\[ {\bf npar} \]

long nPoints

Number of multiple shooting intervals.

Command line argument is -nms and not -nPoints !!

long nvar

Number of variables/equations.

\[ {\bf nobs} \le {\bf nvar}\]

double* par

Parameters.

Vector of dimension:

\[ {\bf npar}\;,\]

where npar is defined in structure Glob.

double* r2

Right-hand side of equality constraints.

Vector of dimension:

\[ {\bf me}\]

double* r3

Right-hand side of inequality constraints.

Vector of dimension:

\[ {\bf mg}\]

double** residues

Current residues.

\[ residues_{i,j}=yPred_{i,j}-yMeasure_{i,j}\]

Therefore matrix of dimension:

\[ {\bf nMeasure}\times{\bf nobs}\]

double** sigma

Standard deviation of data.

Matrix of dimension:

\[ {\bf nMeasure}\times{\bf nobs}\]

string* splineFile

Number of observed quantities Name of spline data-file

double* ua

Condensation of the least squares.

\[||u_a+E_a \Delta s_0 +P_a \Delta p ||={\bf min} \]

double* ue

Condensation of the equality constraints.

\[u_e+E_e \Delta s_0 +P_e \Delta p=0 \]

double* ug

Condensation of the inequality constraints.

\[u_g+E_g \Delta s_0 +P_g \Delta p \ge 0 \]

double* xMeasure

Temporal data points.

Vector of dimension:

\[ {\bf nMeasure} \]

double* y0

Initial value.

Vector of dimension:

\[ {\bf nvar}\]

double** yComp

Predicted trajectory (intergation)

Matrix of dimension:

\[ {\bf nMeasure}\times{\bf nvar}\]

double** yMeasure

Data of the observations.

Matrix of dimension:

\[ {\bf nMeasure}\times{\bf nobs}\]

double** yPred

Predicted observables (intergation)

Matrix of dimension:

\[ {\bf nMeasure}\times{\bf nobs}\]

double** yTry

Initial guesses at mesh points.

Matrix of dimension:

\[ {\bf nPoints} \times {\bf nvar} \]


The documentation for this class was generated from the following file: