Bayesian Filtering Library  Generated from SVN r
Todo List
Class BootstrapFilter< StateVar, MeasVar >
The implementation is very slow for the moment. It would probably be much faster to add a vector<WeightedSample> to the private members of this class.
See also
Pdf
Class ConditionalPdf< Var, CondArg >
Investigate if we can allow. It is for sure that we'll need another class then the std::list to implement this!
See also
Pdf
Class DiscreteConditionalPdf
Check if this is the best way to implement this.
Note
that the name of this class could be better chosen. Something like Discrete-DiscreteConditionalPdf would maybe be more clear (???), but quite long...
See also
ConditionalPdf
Member DiscreteConditionalPdf::DiscreteConditionalPdf (int num_states=1, int num_conditional_arguments=1, int cond_arg_dimensions[]=NULL)
Get cleaner api and implementation
Member Filter< StateVar, MeasVar >::_timestep
Check wether this really belongs here
Member Filter< StateVar, MeasVar >::Filter (const Filter< StateVar, MeasVar > &filt)
Check if we should make a copy of the pdf's too?
Class MCPdf< T >
This class can and should be made far more efficient!!!
Class MCPdf< T >
This class can and should be made far more efficient!!!
Member MCPdf< T >::CumulativePDFGet ()
what's the best way to remove some samples?
Member MCPdf< T >::CumulativePDFGet ()
what's the best way to remove some samples?
Class MeasurementModel< MeasVar, StateVar >
Check if there should be a "model" base class...
Note
Contrary to the system model, this template class has 2 template arguments: this is because of the different nature of the 2 conditional densities $ P ( Z | X ) $ and $ P ( X_k | X_{k-1} ) $ If $ X_{k-1} $ is discrete, then $ X_{k} $ will also be discrete, but a discrete state doesn't automatically imply a discrete measurement (as is proven in ASR!)
Class MixtureBootstrapFilter< StateVar, MeasVar >
The implementation is very slow for the moment. It would probably be much faster to add a vector<WeightedSample> to the private members of this class.
See also
Pdf
Class NonminimalKalmanFilter
Seriously reimplement this class!
Class ParticleFilter< StateVar, MeasVar >
: Actually all particle filters represented by this class are of the "Sequential importance sampling methods" type. Typical of those methods is the so called Proposal density. In theory it would be possible to create Filters using a recursive version of other Monte Carlo methods (eg. MCMC methods), although I am not aware of any of these (due to the increased complexity).
Member Pdf< T >::CovarianceGet () const
extend this more general to n-th order statistic
Member Pdf< T >::CovarianceGet () const
extend this more general to n-th order statistic
Member Pdf< T >::SampleFrom (vector< Sample< T > > &list_samples, const unsigned int num_samples, int method=DEFAULT, void *args=NULL) const
replace the C-call "void * args" by a more object-oriented structure: Perhaps something like virtual Sample * Sample (const int num_samples,class Sampler)
Member Pdf< T >::SampleFrom (vector< Sample< T > > &list_samples, const unsigned int num_samples, int method=DEFAULT, void *args=NULL) const
replace the C-call "void * args" by a more object-oriented structure: Perhaps something like virtual Sample * Sample (const int num_samples,class Sampler)
Class SystemModel< T >
Check if there should be a "model" base class...
Member vector< T, A >::vector (size_type, const array_type &data)
remove this definition because size is not used