#include <PostProcessingWriter.hpp>

Collaboration diagram for PostProcessingWriter< ELEMENT_DIM, SPACE_DIM >:

Public Member Functions
	PostProcessingWriter (AbstractTetrahedralMesh< ELEMENT_DIM, SPACE_DIM > &rMesh, const FileFinder &rDirectory, const std::string &rHdf5FileName, const std::string &rVoltageName="V", hsize_t hdf5DataWriterChunkSize=0)

void	WritePostProcessingFiles ()

	~PostProcessingWriter ()

void	WriteAboveThresholdDepolarisationFile (double threshold)

Private Member Functions
void	WriteVariablesOverTimeAtNodes (std::vector< unsigned > &rNodeIndices)

void	WriteApdMapFile (double repolarisationPercentage, double threshold)

void	WriteUpstrokeTimeMap (double threshold)

void	WriteMaxUpstrokeVelocityMap (double threshold)

void	WriteConductionVelocityMap (unsigned originNode, std::vector< double > distancesFromOriginNode)

void	WriteGenericFileToMeshalyzer (std::vector< std::vector< double > > &rDataPayload, const std::string &rFolder, const std::string &rFileName)

void	WriteOutputDataToHdf5 (const std::vector< std::vector< double > > &rDataPayload, const std::string &rDatasetName, const std::string &rDatasetUnit, const std::string &rUnlimitedVariableName="PaceNumber", const std::string &rUnlimitedVariableUnit="dimensionless")

std::string	ConvertToHdf5FriendlyString (double threshold)

Private Attributes
FileFinder	mDirectory

std::string	mHdf5File

std::string	mVoltageName

Hdf5DataReader *	mpDataReader

PropagationPropertiesCalculator *	mpCalculator

unsigned	mLo

unsigned	mHi

AbstractTetrahedralMesh < ELEMENT_DIM, SPACE_DIM > &	mrMesh

hsize_t	mHdf5DataWriterChunkSize

Friends
class	TestPostProcessingWriter

Detailed Description

template<unsigned ELEMENT_DIM, unsigned SPACE_DIM>
class PostProcessingWriter< ELEMENT_DIM, SPACE_DIM >

Write out physiological parameters at the end of a simulation

APD map
Upstroke time map
Upstroke Velocity map
Conduction Velocity map

N.B. You should only ever have one PostProcessingWriter around at once, as multiple Hdf5Readers (a member variable of this class) seem to cause problems.

Definition at line 57 of file PostProcessingWriter.hpp.

Constructor & Destructor Documentation

template<unsigned ELEMENT_DIM, unsigned SPACE_DIM>

PostProcessingWriter< ELEMENT_DIM, SPACE_DIM >::PostProcessingWriter	(	AbstractTetrahedralMesh< ELEMENT_DIM, SPACE_DIM > &	rMesh,
		const FileFinder &	rDirectory,
		const std::string &	rHdf5FileName,
		const std::string &	rVoltageName = `"V"`,
		hsize_t	hdf5DataWriterChunkSize = `0`
	)

Constructor

Parameters

rMesh	A reference to the mesh used to calculate the distance map to pass to the conduction velocity calculator.
rDirectory	The directory the data is in. The output is written to <directory>/output
rHdf5FileName	The file the data is in.
rVoltageName	(Optional) The name of the variable representing the membrane potential. It is used in the creation of the PropagationPropertiesCalculator object. Defaults to "V".
hdf5DataWriterChunkSize	(Optional) Chunk size and alignment parameter to pass to Hdf5DataWriter

Definition at line 52 of file PostProcessingWriter.cpp.

template<unsigned ELEMENT_DIM, unsigned SPACE_DIM>

PostProcessingWriter< ELEMENT_DIM, SPACE_DIM >::~PostProcessingWriter ( )

Destructor

Definition at line 160 of file PostProcessingWriter.cpp.

Member Function Documentation

template<unsigned ELEMENT_DIM, unsigned SPACE_DIM>

std::string PostProcessingWriter< ELEMENT_DIM, SPACE_DIM >::ConvertToHdf5FriendlyString ( double threshold )

private

Convert a string with numbers in it into alphanumeric plus underscores.

e.g. 20 -> "_20" -20 -> "_minus_20" 30.2 -> "_30pt20" -11.238 -> "_minus_11pt23" (always does decimals to (floor) 2d.p.)

Parameters

threshold A numerical threshold which may contain minuses or a decimal point.

Returns: A string version of the number without minuses or decimal points.

Definition at line 394 of file PostProcessingWriter.cpp.

template<unsigned ELEMENT_DIM, unsigned SPACE_DIM>

void PostProcessingWriter< ELEMENT_DIM, SPACE_DIM >::WriteAboveThresholdDepolarisationFile ( double threshold )

Method for opening a file and writing one row per node: line 1: <number of upstrokes for node 0> <number of above-threshold depolarisations for node 0> line 2: <number of upstrokes for node 1> <number of above-threshold depolarisations for node 1> etc.

For the nodes where the threshold isn't crossed, the 'number of upstrokes' will be 0 (so will the number of above-threshold depolarisations for that node)

Todo:: This method ought to be private and called by the WritePostProcessingFiles method if the user requests for it. This will be possible after modifying the schema and specifying Get and Set methods in HeartConfig to check whether the user wants this file or not

Parameters

threshold - used to signify the upstroke (mV) AND to specify above which voltage value the depolarisations are counted

Definition at line 340 of file PostProcessingWriter.cpp.

template<unsigned ELEMENT_DIM, unsigned SPACE_DIM>

void PostProcessingWriter< ELEMENT_DIM, SPACE_DIM >::WriteApdMapFile	(	double	repolarisationPercentage,
		double	threshold
	)

private

Method for opening an APD map file and writing one row per node line 1: <first APD for node 0> <second APD for node 0> ... line 2: <first APD for node 1> <second APD for node 1> ... etc.

Nodes where there is no APD are respresented by a single 0

Parameters

repolarisationPercentage	eg. 90.0 for APD90
threshold	- Vm used to signify the upstroke (mV)

Definition at line 248 of file PostProcessingWriter.cpp.

template<unsigned ELEMENT_DIM, unsigned SPACE_DIM>

void PostProcessingWriter< ELEMENT_DIM, SPACE_DIM >::WriteConductionVelocityMap	(	unsigned	originNode,
		std::vector< double >	distancesFromOriginNode
	)

private

Write out conduction velocity map from the given node the rest of the mesh:

line 1: <conduction velocity for node 0 and AP 0> <conduction velocity for node 0 and AP 1> ... line 2: <conduction velocity for node 1 and AP 0> <conduction velocity for node 1 and AP 1> ... etc.

Note: the line corresponding to node number originNode will contain ...

Parameters

originNode	- Node to compute the conduction velocity from
distancesFromOriginNode	- Distance map from originNode to all the nodes in the simulation. Typically calculated with DistanceMapCalculator

Definition at line 314 of file PostProcessingWriter.cpp.

template<unsigned ELEMENT_DIM, unsigned SPACE_DIM>

void PostProcessingWriter< ELEMENT_DIM, SPACE_DIM >::WriteGenericFileToMeshalyzer	(	std::vector< std::vector< double > > &	rDataPayload,
		const std::string &	rFolder,
		const std::string &	rFileName
	)

private

Method for opening a file and writing one row per node line 1: <first scalar data for node 0> <second scalar data for node 0> ... line 2: <first scalar data for node 1> <second scalar data for node 1> ... etc.

Parameters

rDataPayload	vector data for each node. Each node's data are represented by a vector of scalars (variable length)
rFolder	subfolder for postprocessing in which to put the data.
rFileName	where to put the data.

Definition at line 466 of file PostProcessingWriter.cpp.

References PetscTools::AmMaster(), PetscTools::AmTopMost(), PetscTools::BeginRoundRobin(), PetscTools::EndRoundRobin(), ChasteBuildInfo::GetProvenanceString(), HeartConfig::Instance(), and OutputFileHandler::OpenOutputFile().

template<unsigned ELEMENT_DIM, unsigned SPACE_DIM>

void PostProcessingWriter< ELEMENT_DIM, SPACE_DIM >::WriteMaxUpstrokeVelocityMap ( double threshold )

private

Write out velocities of each max upstroke for each node:

line 1: <first upstroke velocity for node 0> <second upstroke velocity for node 0> ... line 2: <first upstroke velocity for node 1> <second upstroke velocity for node 1> ... etc.

If there is no upstroke then there will a blank line

Parameters

threshold - Vm used to signify the upstroke (mV)

Definition at line 288 of file PostProcessingWriter.cpp.

template<unsigned ELEMENT_DIM, unsigned SPACE_DIM>

void PostProcessingWriter< ELEMENT_DIM, SPACE_DIM >::WriteOutputDataToHdf5	(	const std::vector< std::vector< double > > &	rDataPayload,
		const std::string &	rDatasetName,
		const std::string &	rDatasetUnit,
		const std::string &	rUnlimitedVariableName = `"PaceNumber"`,
		const std::string &	rUnlimitedVariableUnit = `"dimensionless"`
	)

private

Put the post-processed data into the main HDF5 results file.

Parameters

rDataPayload	The postprocessed quantities
rDatasetName	The name of the quantities
rDatasetUnit	The unit of the quantities
rUnlimitedVariableName	The name of the unlimited variable (defaults to "PaceNumber")
rUnlimitedVariableUnit	The unlimited variable units (defaults to "dimensionless")

Definition at line 167 of file PostProcessingWriter.cpp.

References DistributedVector::Begin(), RelativeTo::ChasteTestOutput, DistributedVectorFactory::CreateDistributedVector(), DistributedVectorFactory::CreateVec(), Hdf5DataWriter::DefineVariable(), PetscTools::Destroy(), and DistributedVector::End().

template<unsigned ELEMENT_DIM, unsigned SPACE_DIM>

void PostProcessingWriter< ELEMENT_DIM, SPACE_DIM >::WritePostProcessingFiles ( )

Write out data files. The data that is written depends on which maps have been requested using either the XML file or HeartConfig

Todo:: #2359 work out how to get rid of this line needed because PseudoEcgCalculator makes its own Hdf5DataReader.

Todo:: #2359 work out how to get rid of these lines

Definition at line 72 of file PostProcessingWriter.cpp.

References DistanceMapCalculator< ELEMENT_DIM, SPACE_DIM >::ComputeDistanceMap(), HeartConfig::GetApdMaps(), HeartConfig::GetConductionVelocityMaps(), HeartConfig::GetMaxUpstrokeVelocityMaps(), HeartConfig::GetNodalTimeTraceRequested(), HeartConfig::GetPseudoEcgElectrodePositions(), HeartConfig::GetUpstrokeTimeMaps(), HeartConfig::Instance(), and PseudoEcgCalculator< ELEMENT_DIM, SPACE_DIM, PROBLEM_DIM >::WritePseudoEcg().

Referenced by AbstractCardiacProblem< ELEMENT_DIM, SPACE_DIM, PROBLEM_DIM >::CloseFilesAndPostProcess().

template<unsigned ELEMENT_DIM, unsigned SPACE_DIM>

void PostProcessingWriter< ELEMENT_DIM, SPACE_DIM >::WriteUpstrokeTimeMap ( double threshold )

private

Write out times of each upstroke for each node:

line 1: <first upstroke time for node 0> <second upstroke time for node 0> ... line 2: <first upstroke time for node 1> <second upstroke time for node 1> ... etc.

If there is no upstroke then there will a blank line

Parameters

threshold - Vm used to signify the upstroke (mV)

Definition at line 262 of file PostProcessingWriter.cpp.

template<unsigned ELEMENT_DIM, unsigned SPACE_DIM>

void PostProcessingWriter< ELEMENT_DIM, SPACE_DIM >::WriteVariablesOverTimeAtNodes ( std::vector< unsigned > & rNodeIndices )

private

Method that extrapolates the output variables over time at specified nodes and output all to file. The use of this method is intended as follows: the user supplies a list of node indices (rNodeIndices). This method outputs the time series at each node in rNodeIndices (one time series per column). The node numbering is referred to the original mesh.

Regardless of the permutation that was used in the simulation, this method will have the same output.

There will be one file per variable contained in the hdf5 file. For each file, the name will be NodalTraces_VARIABLENAME.dat. So for example, in a bidomain simulation with default variable names, output files will be:

NodalTraces_V.dat NodalTraces_Phi_e.dat

Each output file will look like (taking variable V as an example):

Line 1: <V at time 0, Node 0> <V at time 0, Node 1> <V at time 0, Node 2> etc... Line 2: <V at time 1, Node 0> <V at time 1, Node 1> <V at time 1, Node 2> etc... etc...

where Node 0, Node 1, Node 2 are the nodes included in the rIndices vector and "V" is the name of the variable.

Parameters

rNodeIndices the node indices (in the unpermuted mesh) that we want the output for.

Definition at line 419 of file PostProcessingWriter.cpp.

References PetscTools::AmMaster(), HeartConfig::GetOutputUsingOriginalNodeOrdering(), and HeartConfig::Instance().