jpsth: Related Functions and Methods for Joint-PSTHs and Joint...
In STAR: Spike Train Analysis with R

Description Usage Arguments Details Value Note Author(s) References See Also Examples

Some mainly graphical tools to probe interactions between 2 neurons recorded in the presence of a repeated stimulation.

jsd(xRT, yRT, acquisitionWindow, xlab, ylab,
    main, pch = ".", ...)
jpsth(xRT, yRT, xBreaks, yBreaks,
      acquisitionWindow, nbEvtPerBin = 50)
## S3 method for class 'jpsth'
contour(x, xlab, ylab, main, ...)
## S3 method for class 'jpsth'
image(x, xlab, ylab, main, ...)
## S3 method for class 'jpsth'
persp(x, xlab, ylab, main, ...)
jpsth2df(object)

`xRT`	a `repeatedTrain` object whose spike times will appear on the abscissa of the plots.
`yRT`	a `repeatedTrain` object whose spike times will appear on the ordinate of the plots. It must have the same length as `xRT`.
`x, object`	`jpsth` objects.
`xBreaks, yBreaks`	A single number (the bin width) or a vector defining bins boundaries on the X and Y axis. If missing a default is provided.
`acquisitionWindow`	a 2 elements vector specifying the begining and the end of the acquisition. If missing values are obtained using the `floor` of the smallest spike time and the `ceiling` of the largest one.
`nbEvtPerBin`	If both `xBreaks` and `xBreaks` are missing a bin width, `bw`, is computed such that the expected value of the count per cell (2 dimensional bin) would be `nbEvtPerBin` assuming a stationary Poisson discharge for both neurons.
`xlab`	a character (default value supplied). See `plot`.
`ylab`	a character (default value supplied). See `plot`.
`main`	a character (default value supplied). See `plot`.
`pch`	the type of "points" displayed by `jsd`. See `plot`.
`...`	additional arguments passed to `plot` by `jsd` and to respective generic methods by `contour.jpsth`, `image.jpsth` and `persp.jpsth`.

The joint scatter diagram was introduced by Gerstein and Perkel (1972). The joint peristimulus time histogram is a binned version of it (Aertsen et al, 1989). jpsth2df allows the reformating of a jpsth object in order to compute a smooth version of it with gssanova, gssanova0 or gam.

jsd is used for its side effect, a plot is generated and nothing is returned.

jpsth2df returns a data.frame with the following variables: Count, the counts per cell; X, the position of the cell on the X axis; Y, the position of the cell on the Y axis; and attributes: xBreaks, yBreaks, xTotal, yTotal, nbTrials, acquisitionWindow corresponding to the components of its argument with the same name and originalCall corresponding to component call.

jpsth returns a list of class jpsth with the following components:

`counts`	a matrix storing the counts per cell.
`density`	a matrix storing the density in each cell.
`xMids`	a vector containing the X positions of the cells.
`yMids`	a vector containing the Y positions of the cells.
`xBreaks`	a vector containing the bin boundaries of the cells along the X axis.
`yBreaks`	a vector containing the bin boundaries of the cells along the X axis.
`xTotal`	the total number of spikes of the "X" neuron.
`yTotal`	the total number of spikes of the "Y" neuron.
`xFreq`	the mean freqency of the "X" neuron.
`yFreq`	the mean freqency of the "Y" neuron.
`nbTrials`	the number of trials of `xRT` (and `yRT`).
`acquisitionWindow`	the boundaries of the acquisition window.
`call`	the matched call.

I use "joint scatter diagram" for what Gerstein and Perkel (1972) more properly call a "joint peristimulus time scatter diagram".

Christophe Pouzat christophe.pouzat@gmail.com

Gerstein, G. L. and Perkel, D. H. (1972) Mutual temporal relationships among neuronal spike trains. Statistical techniques for display and analysis. Biophys J 12: 453–473. http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1484144

Aertsen, A. M., Gerstein, G. L., Habib, M. K., Palm, G. (1989) Dynamics of neuronal firing correlation: modulation of "effective connectivity". J Neurophysiol 61: 900–917. http://jn.physiology.org/cgi/content/abstract/61/5/900

lockedTrain, plot.lockedTrain, hist.lockedTrain, gsslockedTrain, plot.gsslockedTrain, gsslockedTrain0, plot.gsslockedTrain0, gamlockedTrain, plot.gamlockedTrain, contour, image, persp, attr, attributes

## load e070528citronellal data
data(e070528citronellal)
## plot a jsd with neuron 1 on X and neuron 2 on Y
jsd(e070528citronellal[[1]],e070528citronellal[[2]])
## now make the jpsth
j1.2 <- jpsth(e070528citronellal[[1]],e070528citronellal[[2]])
## make a contour plot
contour(j1.2)
## make an image plot
image(j1.2)
## make a persp plot
persp(j1.2)
## Not run: 
## fit a gss model with interactions
## use a larger bin width for the jpsth
j1.2 <- jpsth(e070528citronellal[[1]],e070528citronellal[[2]],0.2,0.2)
## get a data frame
j1.2DF <- jpsth2df(j1.2)
## To save computation time start analyzing
## just before the stimulation time
j1.2DF <- j1.2DF[j1.2DF$X > 6 & j1.2DF$Y>6,]
gf <- gssanova(Count ~ X*Y, family="poisson", data=j1.2DF,seed=20061001)
## Use the project function of gss to check the significance
## of the interaction term
project(gf2,inc=c("X","Y"))

## End(Not run)
## Not run: 
## fit a gam model assuming no interaction
## get a data frame
j1.2DF <- jpsth2df(j1.2)
fitNoI <- gam(Count ~ s(X,k=100,bs="cr") + s(Y,k=100,bs="cr"),data=j1.2DF,family=poisson())

## End(Not run)