Augments a CDat with a dynamically calculated gene set filter. More...

#include <datapair.h>

Inheritance diagram for Sleipnir::CDatFilter:

Public Member Functions
bool	Attach (const CDataPair &Dat, const CGenes &Genes, CDat::EFilter eFilter, const CDat *pAnswers=NULL)
	Associates the data filter with the given CDat, gene set, and filter type.
bool	Attach (const CDatFilter &Dat, const CGenes &Genes, CDat::EFilter eFilter, const CDat *pAnswers=NULL)
	Associates (overlays0 the data filter with the given pre-existing filter, gene set, and filter type.
bool	Attach (const CDataPair &Dat)
	Associates the data filter with the given CDat.
size_t	GetValues () const
	Returns the number of values taken by the underlying CDataPair.
size_t	GetGene (const std::string &strGene) const
	Returns the index of the requested gene name, or -1 if it does not exist.
std::string	GetGene (size_t iGene) const
	Returns the gene name at the requested index.
size_t	Quantize (float dValue) const
	Discretizes the given value using the quantization bins of the underlying CDataPair.
size_t	Quantize (size_t iY, size_t iX, size_t iZero) const
float &	Get (size_t iY, size_t iX) const
	Returns the (potentially filtered) value at the requested indices.

Detailed Description

Augments a CDat with a dynamically calculated gene set filter.

A filter wraps an underlying CDat with a dynamically calculated filter using a gene set and CDat::EFilter type. A filtered gene pair will act like missing data; unfiltered gene pairs will be retrieved from the underlying CDat. This allows data to be temporarily hidden without modifying the underlying (potentially memory mapped) CDat.

Remarks:: Permanent modifications such as CDat::Set should be performed only on the underlying dataset. Yes, an interface would make this a lot cleaner, but it also makes it a lot slower (losing the ability to inline calls to Get actually has a non-trivial impact on runtime).

See also:: CDat::FilterGenes | CDataFilter

Definition at line 235 of file datapair.h.

Member Function Documentation

bool Sleipnir::CDatFilter::Attach	(	const CDataPair &	Dat,
		const CGenes &	Genes,
		CDat::EFilter	eFilter,
		const CDat *	pAnswers = `NULL`
	)

Associates the data filter with the given CDat, gene set, and filter type.

Parameters:

Dat	CDat to be associated with the overlaying mask.
Genes	Gene set used to filter the data.
eFilter	Way in which to use the given genes to remove gene pairs.
pAnswers	If non-null, answer set to be used for filter types requiring answers (e.g. CDat::EFilterTerm).

Returns:: True if filter was attached successfully.

Remarks:: No calculation occurs during Attach; the gene set and filter type are stored, and calculations are performed dynamically in IsExample. The gene set is not copied, so the given object should not be destroyed until after the filter. If a filter types requiring an answer file is given without an accompanying answer file, the filter won't crash, but results might be a little odd.

Definition at line 582 of file datapair.cpp.

Referenced by Attach().

bool Sleipnir::CDatFilter::Attach	(	const CDatFilter &	Dat,
		const CGenes &	Genes,
		CDat::EFilter	eFilter,
		const CDat *	pAnswers = `NULL`
	)

Associates (overlays0 the data filter with the given pre-existing filter, gene set, and filter type.

Parameters:

Dat	Filter with which this mask will associate (and overlay).
Genes	Gene set used to filter the data.
eFilter	Way in which to use the given genes to remove gene pairs.
pAnswers	If non-null, answer set to be used for filter types requiring answers (e.g. CDat::EFilterTerm).

Returns:: True if filter was attached successfully.

Remarks:: This is a slightly ugly way to allow additive filters without requiring a virtual interface. The CDat class is so fundamental, and CDat::Get calls occur so frequently, that a virtual function call can actually have a noticeable impact on performance.

Definition at line 611 of file datapair.cpp.

References Attach().

bool Sleipnir::CDatFilter::Attach ( const CDataPair & Dat ) [inline]

Associates the data filter with the given CDat.

Parameters: