isi {STAR} | R Documentation |
A utility function to create a vector containing the ith preceding inter spike interval (isi) at a given time.
isi(dataFrame, lag = 1)
dataFrame |
a data.frame typically generated by
mkGLMdf . Should at least contain an event
and a time variable. |
lag |
a strictly positive integer. Set to 1 if the previous isi is required, to 2 is the isi preceding the last one is required, etc... |
Look at the (short) source file for details.
A numeric
vector with the value of the lag
th isi
preceding the time of the corresponding bin center.
Before plugging the result into gssanova
, do not forget
to remove the NA
elements (see the example).
Christophe Pouzat christophe.pouzat@gmail.com
## Not run: ## load e060517spont data set data(e060517spont) ## make a data frame using a 2 ms bin width e060517spontDF <- mkGLMdf(e060517spont,0.002,0,60) ## Keep data relevant to neuron 1 e060517spontDFn1 <- e060517spontDF[e060517spontDF$neuron == "1",] ## get the isi at lag 1 and 2 e060517spontDFn1$isi1 <- isi(e060517spontDFn1,lag=1) e060517spontDFn1$isi2 <- isi(e060517spontDFn1,lag=2) ## keep only defined elements e060517spontDFn1 <- e060517spontDFn1[!is.na(e060517spontDFn1$isi2),] ## split the data set into an "early" and a "late" part e060517spontDFn1e <- e060517spontDFn1[e060517spontDFn1$time <= 30,] e060517spontDFn1l <- e060517spontDFn1[e060517spontDFn1$time > 30,] ## Fit the late part e060517spontDFn1lGF <- gssanova(event ~ lN.1*isi1+isi2, data=e060517spontDFn1l, family="binomial", seed=20061001) ## Time transform the early part and perform goodness of fit tests e060517spont.n1e.tt <- e060517spontDFn1lGF %tt% e060517spontDFn1e e060517spont.n1e.tt summary(e060517spont.n1e.tt) plot(summary(e060517spont.n1e.tt)) ## End(Not run)