Inheritance diagram for SetOfNetworkLegos:

Public Member Functions
	SetOfNetworkLegos (SetOfActiveNetworks &anetSet)
virtual void	compute (ostream &logStream, unsigned int minNumberConditions=1, unsigned int minNumberInteractions=1)
void	computeMaximallySignificantItemsets (const vector< Itemset > &allSignificantItemsets, vector< Itemset > &maximallySignificantItemsets, vector< Itemset > &significantLatticeRoots, vector< Itemset > &significantLatticeLeaves)
	Among all the statistically-significant itemsets, compute those that are more significant than all their ancestors and their descendants.
virtual void	compareNetworkLegosToActiveNetworks ()
	Compare each network lego to the active networks it contributes to.
virtual void	computeTrivialLegos (ostream &logStream)
	Compute trivial legos, i.e., the intersection of all active networks and the portion of each active network unique to it.
virtual void	computeRecoverability (ostream &logStream)
virtual void	computeSignificantLegos (ostream &logStream, unsigned int numRandomTrials=100000, MyNT pvalueThreshold=0.01)
	Compute which NetworkLegos are statistically significant.
virtual void	computeStability (ostream &logStream)
	Compute the stability of statistically-significant NetworkLegos.
virtual void	getSignificantLegos (SetOfActiveNetworks &legos)
	Return the significant NetworkLegos.
void	printBinaryMatrix (ostream &ostr, bool transposed=false) const
void	printClosedLattice (ostream &ostr) const
	Print the edges of the transitively-closed lattice to ostr.
void	printLattice (ostream &ostr) const
	Print the edges of the lattice to ostr.
virtual void	readMappingFromNetworkLegosToActiveNetworks (string nlsToRNsFile)
	Read the mapping that specifies how each network lego contributes to each active network.

Member Function Documentation

void SetOfNetworkLegos::compareNetworkLegosToActiveNetworks ( ) [virtual]

Compare each network lego to the active networks it contributes to.

For each network lego, find the active networks it contributes to. For each such active network, compute the overlap between the network lego and the active networks in terms of the number of nodes, the set similarity (Jaccard's co-efficient) between the two node sets, the two asymmetric set similarities ("one-sided" Jaccard's coefficient) between the two nodes sets, and similar quantities for the edge sets.

Note:: The method assumes that the mapping from networks legos to the active networks they contribute to has already been read from the appropriate input file by a call to SetOfNetworkLegos::readMappingFromNetworkLegosToActiveNetworks().

Reimplemented from SetOfActiveNetworks.

void SetOfNetworkLegos::compute	(	ostream &	logStream,
		unsigned int	minNumberConditions = `1`,
		unsigned int	minNumberInteractions = `1`
	)		`[virtual]`

Compute NetworkLegos

Parameters:

[in]	logStream,an	output stream to log information to.
[in]	numTrials,the	number of random trials to use when computing statistical significance.

void SetOfNetworkLegos::computeMaximallySignificantItemsets	(	const vector< Itemset > &	allSignificantItemsets,
		vector< Itemset > &	maximallySignificantItemsets,
		vector< Itemset > &	significantLatticeRoots,
		vector< Itemset > &	significantLatticeLeaves
	)

Among all the statistically-significant itemsets, compute those that are more significant than all their ancestors and their descendants.

Parameters:

[in]	allSignificantItemsets,a	vector of all statistically-significant itemsets.
[out]	maximallySignificantItemsets,a	vector of all maximally-significant itemsets.
[out]	significantLatticeRoots,a	vector of the roots of the lattice formed by the statistically-significant itemsets.
[out]	significantLatticeLeaves,a	vector of the leaves of the lattice formed by the statistically-significant itemsets.

The method computes the lattice connecting allSignificantItemsets. It sets significantLatticeRoots to be the itemsets at the roots of this lattice and significantLatticeLeaves to be the itemsets at the leaves of this lattice.

An itemset is maximally-significant if it is more statistically significant than all its ancestors and all its descendants. The vector maximallySignificantItemsets contains these itemsets.

void SetOfNetworkLegos::computeStability ( ostream & logStream ) [virtual]

Compute the stability of statistically-significant NetworkLegos.

void SetOfNetworkLegos::printBinaryMatrix	(	ostream &	ostr,
		bool	transposed = `false`
	)		const

Print the binary matrix to ostr

Parameters:

[in] transposed,if true, transpose the rows and columns.

void SetOfNetworkLegos::readMappingFromNetworkLegosToActiveNetworks ( string nlsToRNsFile ) [virtual]

Read the mapping that specifies how each network lego contributes to each active network.

Parameters:

[in] nlsToRNsFile a three-column file specifying the mapping. The file is tab-delimited. The first column is the name of the condition corresponding to the active network. The second column is the identifier of the network lego. The third column is the weight by which the network lego contributes to the active network.

Note:: A typical example of this file is "legos-to-conditions.txt" produced by the inmf.pl script from the Z.txt computed by the INMF algorithm. Thus, this file is a three column representation of Z.txt.

Reimplemented from SetOfActiveNetworks.

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

/home/poirel/src/c++/biorithm/active-networks/network-lego.h
/home/poirel/src/c++/biorithm/active-networks/network-lego.C

Public Member Functions

Member Function Documentation