This function uses CLIP (combination of likelihood profiles) to compute the pooled profile of the posterior after multiple imputation.
CLIP.profile(
obj = NULL,
variable,
data,
which,
firth = TRUE,
weightvar,
control = logistf.control(),
offset = NULL,
from = NULL,
to = NULL,
steps = 101,
legacy = FALSE,
keep = FALSE
)An object of class CLIP.profile with items:
The values of the regression coefficient
The cumulative distribution function of the posterior
The profile of the posterior
An imputations x steps matrix with the values of the completed-data CDFs for each beta
An imputations x steps matrix with the values of the completed-data profiles for each beta
The function call
Either a list of logistf fits (on multiple imputed data sets), or the result
of analysis of a mice (multiply imputed) object using with.mids.
The variable of interest, for which confidence intervals should be computed. If missing, confidence intervals for all variables will be computed.
A list of data set corresponding to the model fits. Can be left blank if obj was obtained with the dataout=TRUE option or if obj was obtained by mice.
Alternatively to variable, the argument which allows to specify the variable to compute the profile for as righthand formula, e.g. which=~X.
If TRUE, applies the Firth correction. Should correspond to the entry in obj.
An optional weighting variable for each observation
control parameters for logistf, usually obtained by logistf.control()
An optional offset variable
Lowest value for the sequence of values for the regression coefficients for which the profile will be computed. Can be left blank.
Highest value for the sequence of values for the regression coefficients for which the profile will be computed. Can be left blank
Number of steps for the sequence of values for the regression coefficients for which the profile will be computed
If TRUE, only R code will be used. Should be avoided.
If TRUE, keeps the profiles for each imputed data sets in the output object.
Georg Heinze und Meinhard Plonar
While CLIP.confint iterates to find those values at which the CDF of the pooled posterior equals the confidence levels, CLIP.profile will evaluate the whole profile, which enables plotting and evaluating the skewness of the combined and the completed-data profiles. The combined and completeddata profiles are available as cumulative distribution function (CDF) or in the scaling of relative profile likelihood (minus twice the likelihood ratio statistic compared to the maximum). Using a plot method, the pooled posterior can also be displayed as a density.
Heinze G, Ploner M, Beyea J (2013). Confidence intervals after multiple imputation: combining profile likelihood information from logistic regressions. Statistics in Medicine, to appear.
#generate data set with NAs
freq=c(5,2,2,7,5,4)
y<-c(rep(1,freq[1]+freq[2]), rep(0,freq[3]+freq[4]), rep(1,freq[5]), rep(0,freq[6]))
x<-c(rep(1,freq[1]), rep(0,freq[2]), rep(1,freq[3]), rep(0,freq[4]), rep(NA,freq[5]),
rep(NA,freq[6]))
toy<-data.frame(x=x,y=y)
# impute data set 5 times
set.seed(169)
toymi<-list(0)
for(i in 1:5){
toymi[[i]]<-toy
y1<-toymi[[i]]$y==1 & is.na(toymi[[i]]$x)
y0<-toymi[[i]]$y==0 & is.na(toymi[[i]]$x)
xnew1<-rbinom(sum(y1),1,freq[1]/(freq[1]+freq[2]))
xnew0<-rbinom(sum(y0),1,freq[3]/(freq[3]+freq[4]))
toymi[[i]]$x[y1==TRUE]<-xnew1
toymi[[i]]$x[y0==TRUE]<-xnew0
}
# logistf analyses of each imputed data set
fit.list<-lapply(1:5, function(X) logistf(data=toymi[[X]], y~x, pl=TRUE))
# CLIP profile
xprof<-CLIP.profile(obj=fit.list, variable="x",data =toymi, keep=TRUE)
plot(xprof)
#plot as CDF
plot(xprof, "cdf")
#plot as density
plot(xprof, "density")
Run the code above in your browser using DataLab