opt.bins: Find the optimal way to bin the response times.
In matsukik/mrsat: Multiple Response Speed Accuracy Tradeoff

Description Usage Arguments Details Value See Also Examples

A function to divide the response time data into a fixed number of small consecutive intervals or bins, in a way that is optimal given the distribution of the reponse time data. This involves finding an optimal offset (or start) of bining, and possily an optimal bin size.

opt.bins(RT, maxRT = 6300, breaksize = 7, 
  startbin = 1, endbin = 16, binsize = 350, 
  type = c("fixed", "free"),
  add.two=FALSE, shift.edge = FALSE)

`RT`	a numeric vector containing the raw response time information
`maxRT`	numeric. Maximum duration of response time above which will be trimmed for the estimation.
`breaksize`	a single number representing the size of bins (in milisecond) used in the intermediate step.
`startbin`	integer.
`endbin`	integer.
`binsize`	a single number corresponding to the initial (and final if `type="fixed"`) window size to be used for binning
`type`	a character string indicating whether to use the window of prespecified width ("fixed") or the widow of data-specific width ("free").
`add.two`	logical. An option for compatibility. If `TRUE`, a magic number 2 is added to the window size when calculating the fit.
`shift.edge`	logical. An option forcompatibility with old codes. If `TRUE`, the timing bins are shifted 100 ms forward. Default to `FALSE`.

This function is meant to be called internally from a wapper function get.bins.

The optimal timing bin size (if window="free") and offset are estimated by fitting the half-rectified sine wave against the histgram of response data.

A list containing:

`maxRT, breaksize, endbin, binsize, type`	same as scpecified in the arguemnts
`peaktime`	a vector of time points representing peak in each time bin
`offset`	difference in time between the offset of the stimulus and the onset of the fastest response.
`fit`	root mean squared distance between the actual distribution of the data and the half-rectified sine wave
`phase`	phase (in ms) for the sign wave function.
`omega`	fitted window size. If `type="fixed"`, the value should be the same as the argument `binsize`
`rt.hist`	an object of class "histogram", with `x=RT`, and `breaks=seq(from=0, to=maxRT, by=breaksize)`
`amp`	amplitude for the sign wave function.

get.bins

data(Auditory_demo)
rt_col <- grep("rt[0-3][0-9]",colnames(Auditory_demo), value=TRUE)
RT <- Auditory_demo[,rt_col]
d.opt <- opt.bins(unlist(na.omit(RT)))
plot(d.opt)