DstarM (version 0.2.2)

chisqFit: Calculate model fit

Description

Calculate model fit

Usage

chisqFit(resObserved, data, DstarM = FALSE, tt = NULL)

Arguments

resObserved

either output from estObserved or a matrix containing custom densities to calculate the fitness for.

data

A dataframe containing data.

DstarM

Logical. Should the DstarM fit measure be calculated or the traditional fit measure?

tt

time grid custom densities where calculated on. Should only be supplied if resOberved is a matrix containing custom densities

Details

This function allows a user to manually calculate a chi-square goodness of fit measure for model densities. This is useful for comparing a traditional analysis and a D*M analysis. For completion, this function can also calculate a D*M fit measure. We do not recommend usage of the D*M measure. While the chi-square fit measure is identical to the value of the optimizer when fitting, the DstarM fit measure is not equal to that of a DstarM analysis. This is because this function calculates the DstarM fit measure on the complete distribution, not on the model distributions, as is done during the optimization.

Examples

# NOT RUN {
tt = seq(0, 5, .1)
pars = c(.8, 2, .5, .5, .5, # condition 1
         .8, 3, .5, .5, .5,  # condition 2
         .8, 4, .5, .5, .5)  # condition 3
pdfND = dbeta(tt, 10, 30)

# simulate data
allDat = simData(n = 3e3, pars = pars, tt = tt, pdfND = pdfND, return.pdf = TRUE)
truePdf = allDat$pdfUnnormalized
dat = allDat$dat
chisqFit(resObserved = truePdf, data = dat, tt = tt)
# }
# NOT RUN {
# estimate it
define restriction matrix
restr = matrix(1:5, 5, 3)
restr[2, 2:3] = 6:7 # allow drift rates to differ
# fix parameters for speed up
fixed = matrix(c("z1", "a1 / 2", 'sz1', .5, 'sv1', .5), 2, 3)
resD = estDstarM(data = dat, tt = tt, restr = restr, fixed = fixed,
                 Optim = list(parallelType = 1))
resN = estND(resD, Optim = list(parallelType = 1))

resO = estObserved(resD, resN, data = dat)
resO$fit # proper fit
# }