These M-files implement the analysis procedures discussed in chapter 9 of “Methods in Neuronal Modeling”. In addition, many theoretical results described there are also implemented in the form of M-files, allowing a direct comparison between simulations and theory. References to the literature are provided in the M-files themselves or through the Matlab help utility. This includes the following routines:
Data Analysis
countprob: returns the probability density of the spike count in a fixed time interval.
dispersion: returns the variance in the spike count as a function of the mean spike count for a range of time intervals.
isidist: computes the interspike interval distribution and returns its mean and coefficient of variation.
psautostim: estimates the power spectrum and autocorrelation function of a random input stimulus.
psautospk: estimates the power spectrum and autocorrelation function of a spike train.
roc: computes from two spike count probability distributions the ROC curve characterizing the discrimination performance of an ideal observer based on these distributions.
stimest: estimates a random stimulus from a spike train by Wiener-Kolmogorov filtering.
wiener1: estimates the transfer function characterizing the transformation between a random stimulus and output spike train.
Theoretical Routines
buttfilt*: plots the transfer function of Butterworth low-pass filter.
fithifref: returns the F-I curve of a perfect integrate-and-fire neuron with refractory period.
fithlifref: returns the F-I curve of a leaky integrate-and-fire neuron with refractory period.
gaussroc: returns the ROC curve characterizing the discrimination performance of an ideal observer between two Gaussian probability distributions.
gfitheor: returns the ISI distribution of an integrate-and-fire neuron which sums n_th Poisson inputs to reach threshold (this also corresponds to the ISI distribution of an integrate-and-fire neuron with gamma distributed random thereshold of order n_th).
gpstheor: returns the power spectrum and autocorrelation function of an integrate-and-fire neuron which sums n_th Poisson inputs to reach threshold.
lgnfilt*: returns the transfer function of an LGN relay cell. neymtacnt: returns the probability density of a Neyman type A spike count distribution.
poisscnt: returns the probability density of a Poisson spike count distribution.
prefdisp: returns the variance in spike count as a function of mean spike count for a Poisson neuron with refractory period.
sndpdisp: returns the approximate value of the variance in spike count as function of the mean spike count for a doubly stochastic Poisson process where the driving rate is a shot noise with a square impulse response.
stesttheor: returns the fraction of a white Gaussian signal encoded in two half-wave rectifying Poisson neurons.
The M-files marked with an asterisk (*) are mainly convienent to set up simulation models.