63 for (
int i=
tik_.beginRows(); i <
tik_.endRows(); i++)
66 for (
int k=
tik_.beginCols(); k<
tik_.endCols(); k++)
In this file we define the interface base class for learners.
Type const maxElt(int &row, int &col) const
UnaryOperator< IsInfiniteOp< Type >, Derived > isInfinite() const
CVectorXi ziPred_
The predicted class label.
IMixtureLearner(int nbSample, int nbCluster)
Constructor.
virtual void mapStep()
Compute ziPred using the Map estimate.
Clust::modelState state_
state of the model
virtual ~IMixtureLearner()
destructor
CArrayXX tik_
The tik probabilities.
virtual Real lnComponentProbability(int i, int k) const =0
IMixtureStatModel(int nbSample, int nbCluster)
Constructor.
CPointX pk_
The proportions of each mixtures.
Arrays::SumOp< Lhs, Rhs >::result_type sum(Lhs const &lhs, Rhs const &rhs)
convenience function for summing two arrays
double Real
STK fundamental type of Real values.
hidden::SliceVisitorSelector< Derived, hidden::MaxVisitor, Arrays::by_col_ >::type_result max(Derived const &A)
If A is a row-vector or a column-vector then the function will return the usual maximal value of the ...
The namespace Clust enclose all the enum and utilities functions needed by the Clustering project.
The namespace STK is the main domain space of the Statistical ToolKit project.
CArrayPoint< Real, UnknownSize, Arrays::by_col_ > CPointX