---

StatisticsOutput Class Reference

#include <amygdala/statisticsoutput.h>

Inheritance diagram for StatisticsOutput:

List of all members.

Public Methods
	StatisticsOutput ()
virtual	~StatisticsOutput ()
virtual void	OutputEvent (AmIdInt neuronId, AmTimeInt eventTime)
virtual void	OutputCache ()
void	SetStepSize (AmTimeInt step)
AmTimeInt	GetStepSize () const
void	ClearHistory ()
vector< unsigned int > &	Histogram ()
vector< unsigned int > &	Histogram (AmIdInt neuronId)
vector< unsigned int >	Histogram (AmTimeInt start, AmTimeInt end)
vector< unsigned int >	Histogram (AmIdInt neuronId, AmTimeInt start, AmTimeInt end)
float	PeakSpikeRate ()
float	PeakSpikeRate (AmIdInt neuronId)
AmTimeInt	PeakRateTime ()
AmTimeInt	PeakRateTime (AmIdInt neuronId)
AmIdInt	PeakNeuron ()
AmIdInt	MostActiveNeuron ()
unsigned int	TotalOutputSpikes ()
unsigned int	TotalOutputSpikes (AmIdInt neuronId)
float	MeanSpikeRate ()
float	MeanSpikeRate (AmIdInt neuronId)
void	LogSpikeTimes (string filename, AmTimeInt start=0, AmTimeInt end=~0)
void	AddTrace (AmIdInt nId)
void	CloseLog ()
Protected Methods
void	Log (AmIdInt neuronId, AmTimeInt eventTime)
Protected Attributes
hash_map< AmIdInt, vector< AmTimeInt > >	outputHistory
hash_map< AmIdInt, vector< unsigned int > >	histogram
vector< unsigned int >	combinedHistogram
AmIdInt	maxPeakId
AmTimeInt	maxPeakTime
AmIdInt	mostActiveId
unsigned int	mostActiveCount
float	combinedPeakRate
AmTimeInt	combinedPeakTime
float	meanRate
unsigned int	totalSpikeCount
AmTimeInt	stepSize
AmTimeInt	beginTime
AmTimeInt	calcTime
AmTimeInt	lastCalcTime
set< AmIdInt >	traceNeurons
bool	logging
AmTimeInt	logStart
AmTimeInt	logEnd
FILE *	logFd

---

Detailed Description

Provides statistics and logging facilities to debug networks.

See also:

SpikeOutput

Author:

Matt Grover , Rüdiger Koch

---

Member Function Documentation

void StatisticsOutput::AddTrace (  AmIdInt    nId )

Select a neuron for logging. This function must be called repeatedly to select multiple neurons. Calling this function the first time results in turning off logging for all other neurons

void StatisticsOutput::ClearHistory (    )

Clear out the raw data. This may be done if the simulation is run for a period of time and then resumed after doing some analysis of the results. ClearHistory() can be called before resuming if the old data is not wanted for the next round of analysis.

void StatisticsOutput::CloseLog (    )

Close a the file descriptor if open

AmTimeInt StatisticsOutput::GetStepSize (    )  const [inline]

Get the time step size. (ms)

vector<unsigned int> StatisticsOutput::Histogram (  AmIdInt    neuronId, AmTimeInt    start, AmTimeInt    end )

Generate a histogram detailing the activity of neuronId between start and end times. Times are measured in ms.

vector<unsigned int> StatisticsOutput::Histogram (  AmTimeInt    start, AmTimeInt    end )

Generate a histogram detailing the combined output activity for all output neurons between start and end times. Times are measured in ms.

vector<unsigned int>& StatisticsOutput::Histogram (  AmIdInt    neuronId )

Generate a histogram detailing the activity of neuronId during the life of the simulation.

vector<unsigned int>& StatisticsOutput::Histogram (    )

Generate a histogram detailing the combined output activity for all output neurons during the life of the simulation.

void StatisticsOutput::Log (  AmIdInt    neuronId, AmTimeInt    eventTime )  [protected]

write logging info to the log file

void StatisticsOutput::LogSpikeTimes (  string    filename, AmTimeInt    start = 0, AmTimeInt    end = ~0 )

Turn on logging Parameters: filename  The logfile start  simTime when logging starts end  simTime when logging ends

float StatisticsOutput::MeanSpikeRate (  AmIdInt    neuronId )

Find the mean spike rate for neuron neuronId. (spikes/sec)

float StatisticsOutput::MeanSpikeRate (    )

Find the mean spike rate for all output neurons. (spikes/sec)

AmIdInt StatisticsOutput::MostActiveNeuron (    )

Find the neuron with the highest number of output spikes overall.

virtual void StatisticsOutput::OutputCache (    )  [inline, virtual]

If output is turned on in caching mode, this function will be called every time the cache is emptied. Arguments TBD. Implements SpikeOutput.

virtual void StatisticsOutput::OutputEvent (  AmIdInt    neuronId, AmTimeInt    eventTime )  [virtual]

Raise an output event. If output is turned on in streaming mode, this function will be called every time an output neuron spikes. Implements SpikeOutput.

AmIdInt StatisticsOutput::PeakNeuron (    )

Find the neuron with the highest peak spike rate.

AmTimeInt StatisticsOutput::PeakRateTime (  AmIdInt    neuronId )

Find the time of the peak spike rate for neuronId. (ms)

AmTimeInt StatisticsOutput::PeakRateTime (    )

Find the time of the peak spike rate for the combined output of all output neurons. (ms)

float StatisticsOutput::PeakSpikeRate (  AmIdInt    neuronId )

Find the peak spike rate for neuronId. (spikes/sec)

float StatisticsOutput::PeakSpikeRate (    )

Find the peak spike rate for the combined output of all output neurons. (spikes/sec)

void StatisticsOutput::SetStepSize (  AmTimeInt    step )

Set the time step size (ms) to be used for calculating spike rates. This also determines the width of the elements for the histograms.

unsigned int StatisticsOutput::TotalOutputSpikes (  AmIdInt    neuronId )

Find the total number of spikes for neuron neuronId.

unsigned int StatisticsOutput::TotalOutputSpikes (    )

Find the total number of spikes for all output neurons.

---

Member Data Documentation

AmTimeInt StatisticsOutput::beginTime [protected]

Time that statistics collection began (can be reset)

AmTimeInt StatisticsOutput::calcTime [protected]

Time of last reported output spike (us)

hash_map< AmIdInt, vector<unsigned int> > StatisticsOutput::histogram [protected]

The histograms can be filled either during the simulation inside OutputEvent() or later on from the contents of outputHistory.

AmTimeInt StatisticsOutput::lastCalcTime [protected]

Time that last calculation was done (us)

AmTimeInt StatisticsOutput::stepSize [protected]

histogram step size (ms)

set<AmIdInt> StatisticsOutput::traceNeurons [protected]

The set of neurons that get traced. If empty all neurons get traced

---

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

• statisticsoutput.h

---