src/averaging.c File Reference

#include "averaging.h"
#include "clustering.h"
#include "eeg.h"
#include "linalg.h"

Go to the source code of this file.

Functions
Array *	average_example (const Array *data, uint idx[2], double weights[2], OptArgList *optargs)
	example function to pass as SignalAverageFunction doing a pointwise average.
Array *	average_warpmarkers (const Array *data, uint idx[2], double weights[2], OptArgList *optargs)
EEG *	eeg_simple_average (const EEG *eeg)
Array *	hierarchical_average (const Array *data, const Array *distmat, SignalAverageFunction avgfct, OptArgList *optargs)
	Calculate a hierarchical average based on a cluster-analysis.

---

Detailed Description

STATUS: stable Averaging functions.

Definition in file averaging.c.

---

Function Documentation

Array* average_example	(	const Array *	data,
		uint	idx[2],
		double	weights[2],
		OptArgList *	optargs
	)

example function to pass as SignalAverageFunction doing a pointwise average.

This function is only for testing purposes. The "hierarchically" computed pointwise average should be the same as the normal average.

Parameters:

	input	data (N x C x n)
	index	average data[idx[0]] and data[idx[1]]
	weights	number of trials "behind" each average
	optional	arguments

Returns:

the average

Definition at line 40 of file averaging.c.

Array* average_warpmarkers	(	const Array *	data,
		uint	idx[2],
		double	weights[2],
		OptArgList *	optargs
	)

Definition at line 61 of file averaging.c.

EEG* eeg_simple_average

(

const EEG *

eeg

)

Calculate a simple, pointwise average across trials

Parameters:

eeg

input

Returns:

freshly allocate EEG-struct

Definition at line 189 of file averaging.c.

Array* hierarchical_average	(	const Array *	data,
		const Array *	distmat,
		SignalAverageFunction	avgfct,
		OptArgList *	optargs
	)

Calculate a hierarchical average based on a cluster-analysis.

The data is average based on the first dimension of the data. I.e., the output is C x n dimensional.

A cluster analysis is carried out using the distance-matrix and the resulting dendrogram is followed to average two trials at a time:

The averaging uses a callback-function to calculate the average of two signals at a time. This function must be passed by the use of a SignalAverageFunction .

Parameters:

	data	2D or 3D double-array containing: N trials (dimension 1) of dimensionality C (dimension 2); e.g. channels and n time-points (dimension 3) if the data is 2D, we assume C=1
	distmat	a distance matrix (N x N) given distances between trials in data
	avgfct	a callback-function to calculate the average of two Cxn signals at a time
	optargs	may contain: Directly used by this function: linkage=void* rule for building the dendrogram, default=dgram_dist_completelinkage progress=void* progress-bar function, called every now and then. You might want to pass additional optargs for the SignalAverageFunction and the linkage function

Returns:

pointer to final average (Cxn)

Definition at line 92 of file averaging.c.