Scine::Molassembler::Hashes Namespace Reference

Classes and methods to compute hashes of atom environments. More...

Data Structures
struct	BondInformation
	Information pertinent to a singular bond hash. More...
struct	LookupFunctor
	A functor for getting an atom's hash from a captured list of hashes. More...

Typedefs
using	WideHashType = boost::multiprecision::uint128_t
using	HashType = std::uint64_t

Functions
WideHashType	hash (AtomEnvironmentComponents bitmask, Utils::ElementType elementType, const std::vector< BondInformation > &sortedBonds, const boost::optional< Shapes::Shape > &shapeOptional, const boost::optional< unsigned > &assignedOptional)
	Convolutes the atom's element type and bonds into an unsigned integer. More...
std::vector< BondInformation >	gatherBonds (const PrivateGraph &inner, const StereopermutatorList &stereopermutators, AtomEnvironmentComponents componentsBitmask, AtomIndex i)
	Collects bond information for hash() input. More...
WideHashType	atomEnvironment (const PrivateGraph &inner, boost::optional< const StereopermutatorList & > stereopermutators, AtomEnvironmentComponents bitmask, AtomIndex i)
	Calculate the hash for a particular atom index. More...
std::vector< WideHashType >	generate (const PrivateGraph &inner, boost::optional< const StereopermutatorList & > stereopermutatorsOption, AtomEnvironmentComponents bitmask)
	Generates the hashes for every atom in a molecule's components. More...
bool	identityCompare (const PrivateGraph &aGraph, const StereopermutatorList &aStereopermutators, const PrivateGraph &bGraph, const StereopermutatorList &bStereopermutators, AtomEnvironmentComponents componentBitmask)
	Compare two molecules with an identity index mapping. More...
std::tuple< std::vector < HashType >, std::vector < HashType >, HashType >	narrow (const std::vector< WideHashType > &a, const std::vector< WideHashType > &b)
	Re-enumerates the hashes in two generated hash lists. More...
std::pair< std::vector < HashType >, std::vector < HashType >>	generate (const PrivateGraph &aGraph, const StereopermutatorList &aStereopermutators, const PrivateGraph &bGraph, const StereopermutatorList &bStereopermutators, AtomEnvironmentComponents bitmask)
	Generates hashes for two molecules' components. More...

Detailed Description

Classes and methods to compute hashes of atom environments.

Function Documentation

WideHashType Scine::Molassembler::Hashes::atomEnvironment	(	const PrivateGraph &	inner,
		boost::optional< const StereopermutatorList & >	stereopermutators,
		AtomEnvironmentComponents	bitmask,
		AtomIndex	i
	)

Calculate the hash for a particular atom index.

Complexity \(\Theta(1)\)

std::vector<BondInformation> Scine::Molassembler::Hashes::gatherBonds	(	const PrivateGraph &	inner,
		const StereopermutatorList &	stereopermutators,
		AtomEnvironmentComponents	componentsBitmask,
		AtomIndex	i
	)

Collects bond information for hash() input.

Complexity \(\Theta(1)\) since the out-degree of atoms is considered a fixed small number because molecules are sparse.

std::vector<WideHashType> Scine::Molassembler::Hashes::generate	(	const PrivateGraph &	inner,
		boost::optional< const StereopermutatorList & >	stereopermutatorsOption,
		AtomEnvironmentComponents	bitmask
	)

Generates the hashes for every atom in a molecule's components.

Complexity \(\Theta(N)\)

std::pair< std::vector<HashType>, std::vector<HashType>> Scine::Molassembler::Hashes::generate	(	const PrivateGraph &	aGraph,
		const StereopermutatorList &	aStereopermutators,
		const PrivateGraph &	bGraph,
		const StereopermutatorList &	bStereopermutators,
		AtomEnvironmentComponents	bitmask
	)

Generates hashes for two molecules' components.

Generate hashes for two molecules. Basis for comparing vertices in an isomorphism algorithm.

Complexity \(\Theta(N)\)

WideHashType Scine::Molassembler::Hashes::hash	(	AtomEnvironmentComponents	bitmask,
		Utils::ElementType	elementType,
		const std::vector< BondInformation > &	sortedBonds,
		const boost::optional< Shapes::Shape > &	shapeOptional,
		const boost::optional< unsigned > &	assignedOptional
	)

Convolutes the atom's element type and bonds into an unsigned integer.

At current, this is a bijective mapping and has zero probability of collisions.

Complexity \(\Theta(1)\)

bool Scine::Molassembler::Hashes::identityCompare	(	const PrivateGraph &	aGraph,
		const StereopermutatorList &	aStereopermutators,
		const PrivateGraph &	bGraph,
		const StereopermutatorList &	bStereopermutators,
		AtomEnvironmentComponents	componentBitmask
	)

Compare two molecules with an identity index mapping.

Complexity \(O(N)\)

std::tuple< std::vector<HashType>, std::vector<HashType>, HashType> Scine::Molassembler::Hashes::narrow	(	const std::vector< WideHashType > &	a,
		const std::vector< WideHashType > &	b
	)

Re-enumerates the hashes in two generated hash lists.

Although the hash space requires a wide integer type, the number of different atom environments in molecules will not require a wide integer type. We can therefore remap the wide hashes to an enumerated smaller integer type.

Complexity \(\Theta(N)\)