IMP::em::FitRestraint Class Reference

Detailed Description

Calculate score based on fit to EM map.

Inheritance diagram for IMP::em::FitRestraint:


Public Member Functions
	FitRestraint (Particles ps, DensityMap *em_map, FloatKey radius_key=IMP::core::XYZR::get_default_radius_key(), FloatKey weight_key=IMP::atom::Mass::get_mass_key(), float scale=1, bool special_treatment_of_particles_outside_of_density=true)
	Constructor.
ContainersTemp	get_input_containers () const
ParticlesTemp	get_input_particles () const
ParticlesList	get_interacting_particles () const
SampledDensityMap *	get_model_dens_map ()
virtual std::string	get_type_name () const
virtual ::IMP::VersionInfo	get_version_info () const
virtual double	unprotected_evaluate (DerivativeAccumulator *accum) const
Friends
template<class T >
void	IMP::internal::unref (T *)

IMP::em::FitRestraint::FitRestraint	(	Particles	ps,
		DensityMap *	em_map,
		FloatKey	radius_key = `IMP::core::XYZR::get_default_radius_key()`,
		FloatKey	weight_key = `IMP::atom::Mass::get_mass_key()`,
		float	scale = `1`,
		bool	special_treatment_of_particles_outside_of_density = `true`
	)

Constructor.

Parameters:

`[in]`	ps	The particles participating in the fitting score
`[in]`	em_map	The density map used in the fitting score
`[in]`	radius_key	the name of the radius attribute of the particles
`[in]`	weight_key	the name of the weight attribute of the particles
`[in]`	scale	multiply the fitting restraint score and derivatives by this value
`[in]`	special_treatment_of_particles_outside_of_density	If more than 80% of the particles are outside of the density push it back using upper-bound harmonic

Note:: In many optimization senarios particles are can be found outside of the density. When all particles are outside of the density the cross-correlation score is zero and the derivatives are meaningless. To handle these cases we guide the particles back into the density by using a simple distance restraint between the centroids of the density and the particles. Once the particles are back (at least partly) in the density, the CC score is back on. To smooth the score, we start considering centroids distance once 80% of the particles. This option is still experimental and should be used in caution.

SampledDensityMap* IMP::em::FitRestraint::get_model_dens_map ( )

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