R/msmjson.R
In nskourlis/MSMplus: Prepare MSMplus input files and locally deploys the MSMplus app
Defines functions
msmjson
Documented in
msmjson
#' @title msmjson
#' @description Function msmjson can receive models from msm package.
#' It then uses functions from msm package internally to estimate multi-state model
#' measures such as transition probabilities, transition intensity rates, length of stay, probability of visiting each state
#' probability each state is next, sojoutn times and confidence intervals of the
#' estimations. Function flexjson then take these results and reshapes them so that they can
#' be fed to MSMplus properly as a json file.
#' @param msm.model A fitted multi-state model, as returned by msm.
#' @param vartime The time points to estimate the transition probabilities for, by default one unit.
#' @param mat.init A transition intensity matrix. Default: NULL
#' @param totlos Estimate the total length of stay in each state for the time points
#' specified by vartime argument, Default: "FALSE"
#' @param visit Estimate the probability of visiting each state for the time points
#' specified by vartime argument, Default: "FALSE"
#' @param sojourn Estimate the mean sojourn time in each state, Default: "FALSE"
#' @param pnext Estimate the probability that each state is next, Default: "FALSE"
#' @param efpt Estimate the expected first passage time by each state, Default: "FALSE"
#' @param envisits Estimate the number of visits in each state , Default: "FALSE"
#' @param ci.json Estimate confidence intervals, Default: 'normal'
#' @param cl.json Specify confidence level, Default: 0.95
#' @param B.json Number of simulations from the normal asymptotic distribution used to calculate variances.
#' Decrease for greater speed at the expense of accuracy, Default: 50
#' @param cores.json Number of cores to use for bootstrapping using parallel processing. See boot.msm for more details., Default: NULL
#' @param piecewise.times.json Times at which piecewise-constant intensities change.
#' See pmatrix.piecewise.msm for how to specify this. This is only required for time-inhomogeneous models
#' specified using explicit time-dependent covariates, and should not be used for models specified using "pci"., Default: NULL
#' @param piecewise.covariates.json Covariates on which the piecewise-constant intensities depend. See pmatrix.piecewise.msm for how to specify this., Default: NULL
#' @param num.integ.json Use numerical integration instead of analytic solution (see below).
#' @param covariates_list The user can specify different covariate patterns
#' for which predictions will be made, Default: list()
#' @param jsonpath specify the path of the folder that the json file should be saved, Default: "" saves the json file to the current working directory
#' @param name Specify the name of the output json file, Default: 'predictions.json'
#' @return returns a list of objects: the time variable
#' the number of covariate patterns, the names of covariate patterns, the transition matrix,
#' the number of transitions, the transition probabilities, transition intensity rates, length of stay, probability of visiting each state
#' probability each state is next, sojoutn times and confidence intervals of the
#' estimations(if estimated).
#' @details DETAILS
#' @examples
#' \dontrun{
#'
#' #Example
#'
#' # Multi-state model analysis: Using msmjson function together with msm package
#'
#' library("MSMplus")
#' library("msm")
#' library("dplyr")
#'
#' options(scipen = 999,"digits"=10)
#'
#' head(cav)
#'
#' ### Renaming variable PTNUM to id
#'
#' cav$id=cav$PTNUM
#'
#' ### Defining the transition matrix
#'
#' tmat=matrix(NA,nrow=4,ncol=4)
#' tmat[1,2]=1; tmat[1,4]=2; tmat[2,1]=3; tmat[2,3]=4
#' tmat[2,4]=5; tmat[3,2]=6; tmat[3,4]=7
#'
#' ### We can now call function msboxes_R
#'
#' results3_days=MSMplus::msboxes_R(data=cav,id= cav$id, yb=c(0.3,0.5,0.6,0.75), msm=TRUE,
#' xb=c(0.5,0.2,0.7,0.3),boxwidth=0.1,boxheight=0.1,
#' tmatrix= tmat, vartime=seq(0,10,by=1),scale=1,
#' jsonpath="", name="msboxes_cav_R.json" )
#'
#'
#' ### Defining the transition matrix with initial values under an initial assumption
#'
#' #0 for transitions not allowed, initial values for rest of transitions under a rationale ##
#'
#' Q<- rbind(c(0,0.25,0,0.25),c(0.166,0,0.166,0.166),c(0,0.25,0,0.25),c(0,0,0,0))
#'
#'
#' ### Getting initial Q matrix in a default way- Feed the hand made matrix
#'
#' q.crude<- crudeinits.msm(state~years, id,data=cav, qmatrix=Q)
#'
#'
#' ### Apply the msm model
#'
#' cavsex.msm<- msm(state~years, covariates=~sex, id,data=cav,qmatrix=q.crude, deathexact = 4, control=list(trace=1,REPORT=1))
#' summary(cavsex.msm)
#'
#'
#' ### Prediction for different covariate patterns (males and females)
#'
#' results <- MSMplus::msmjson(msm.model=cavsex.msm, vartime=seq(1,3,1), mat.init=q.crude,
#' totlos=TRUE, visit=TRUE, sojourn=TRUE, pnext=TRUE, efpt=TRUE, envisits=TRUE,
#' ci.json="normal", cl.json=0.95, B.json=100,
#' cores.json=NULL,piecewise.times.json=NULL, piecewise.covariates.json=NULL,num.integ.json=FALSE,
#' covariates_list=list(list(sex = 1),list(sex = 0)),
#' jsonpath="",
#' name="predictions_cav_R.json" )
#'
#' }
#' @seealso
#' \code{\link[stringi]{stri_sort}}
#' @rdname msmjson
#' @export
#' @importFrom stringi stri_sort
msmjson <- function(msm.model, vartime=seq(1,1,1), mat.init,
totlos=FALSE, visit=FALSE, sojourn=FALSE, pnext=FALSE, efpt=FALSE, envisits=FALSE,
ci.json="normal", cl.json=0.95, B.json=50,
cores.json=NULL,piecewise.times.json=NULL, piecewise.covariates.json=NULL,num.integ.json=FALSE,
covariates_list=list(), jsonpath="",name="predictions_R.json" ) {
### Default for jsonpath is user's home directory ####
options(scipen = 999,"digits"=10)
#
# if (is.null(covariate_patterns)) {covariates_list=list()}
#
# else if ( !is.null(covariate_patterns) ) {
# covariates_list=list()
# for (i in 1:nrow(covariate_patterns)) {covariates_list[[i]]=covariate_patterns[i,]}
# }
################################################################
nstates=ncol(mat.init)
p=1
trmat=matrix(nrow=nstates, ncol=nstates, NA)
mat=mat.init
mat[is.na(mat)] <- 0
ntransitions=length(mat[which(mat>0)])
nstates=ncol(mat)
states=c(seq(1:nstates))
trmat=matrix(nrow=nstates, ncol=nstates, NA)
for (i in 1:nstates) {
for (k in 1:nstates) {
if (mat[i,k]>0) {
trmat[i,k]=as.integer(p)
p=p+1
}
else if (mat[i,k]<=0) {
trmat[i,k]=NA
}
}
}
tmatlist=list()
tmatlist[[1]]=trmat
tr_start_state=vector()
for (k in 1:ntransitions) {
tr_start_state[k]=which(trmat == k, arr.ind = TRUE)[1,1]
}
tr_end_state=vector()
for (k in 1:ntransitions) {
tr_end_state[k]=which(trmat == k, arr.ind = TRUE)[1,2]
}
state_kind=vector()
for (i in 1:nstates) {
if (length(which(mat[,i]==0))==nrow(mat)) {state_kind[i]="Starting"}
else if (length(which(mat[i,]==0))==ncol(mat)) {state_kind[i]="Absorbing"}
else {state_kind[i]="Intermediate"}
}
st_states=which(state_kind=="Starting")
na_states=which(state_kind=="Intermediate")
ab_states=which(state_kind=="Absorbing")
intermediate_list=list()
intermediate_list[[1]]=unique(tr_start_state)
Ntransitions=length(mat[which(mat>0)])
###################################################################
# library("msm")
# library("stringi")
# library("RJSONIO")
pred=list()
#if (length(vartime)==1) {vartime=t(vartime)}
### First case, overall predictions
if (length(covariates_list) == 0) {
################################################################################################################
### Create all the lists to be output even if they will remain vacant ###
loslist=list()
loslist_lci=list()
loslist_uci=list()
sojlist=list()
sojlist_lci=list()
sojlist_uci=list()
nextlist=list()
nextlist_lci=list()
nextlist_uci=list()
firstlist=list()
firstlist_lci=list()
firstlist_uci=list()
numberlist=list()
numberlist_lci=list()
numberlist_uci=list()
vlist=list()
vlist_lci=list()
vlist_uci=list()
trans_lci=vector()
trans_uci=vector()
#################################################################################################################
###Probability ###
mat=mat.init
mat[is.na(mat)] <- 0
ntransitions=length(mat[which(mat>0)])
nstates=ncol(mat)
states=c(seq(1:nstates))
##################### Main estimates#####################################################################
p_array= array(dim=c(length(vartime),nstates+1,nstates),NA)
p_array_lci= array(dim=c(length(vartime),nstates+1,nstates),NA)
p_array_uci= array(dim=c(length(vartime),nstates+1,nstates),NA)
for (i in 1:length(vartime)) {
p.msm <- pmatrix.msm(msm.model,t=vartime[i], t1=0, ci=ci.json , cl=cl.json, B=B.json,
cores=cores.json )
if (ci.json=="none" ) {
p_array[i,,]=t(cbind(p.msm,vartime=vartime[i] ))
}
if (ci.json!="none" ) {
p_array[i,,]=t(cbind(p.msm$estimates,vartime=vartime[i] ))
p_array_lci[i,,]=t(cbind(p.msm$L,vartime=vartime[i] ))
p_array_uci[i,,]=t(cbind(p.msm$U,vartime=vartime[i] ))
}
}
plist=list()
for (k in states) {
plist[[k]]=p_array[,,k]
plist[[k]]=as.data.frame(plist[[k]])
}
if (length(vartime)==1) {
for (k in states) {
plist[[k]]=as.data.frame(t(plist[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(plist[[k]])[j]=paste0("P_",k,"_to_",j)
}
colnames(plist[[k]])[nstates+1] ="vartime"
}
plist
plist_lci=list()
plist_uci=list()
if (ci.json!="none" ) {
##################### Probabilities lci #####################################################################
for (k in states) {
plist_lci[[k]]=p_array_lci[,,k]
plist_lci[[k]]=as.data.frame(plist_lci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
plist_lci[[k]]=as.data.frame(t(plist_lci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(plist_lci[[k]])[j]=paste0("P_",k,"_to_",j,"_lci")
}
colnames(plist_lci[[k]])[nstates+1] ="vartime"
}
plist_lci
##################### Probabilities uci #####################################################################
plist_uci=list()
for (k in states) {
plist_uci[[k]]=p_array_uci[,,k]
plist_uci[[k]]=as.data.frame(plist_uci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
plist_uci[[k]]=as.data.frame(t(plist_uci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(plist_uci[[k]])[j]=paste0("P_",k,"_to_",j,"_uci")
}
colnames(plist_uci[[k]])[nstates+1] ="vartime"
}
plist_uci
}
#################################################################################################################
###Transitions values for the period we want (of course they are constant)###
##################### Main estimates#####################################################################
qmsm=qmatrix.msm(msm.model,ci=ci.json , cl=cl.json,
B=B.json, cores=cores.json)
if (ci.json!="none" ) {
trans=matrix(nrow=length(vartime), ncol=length(qmsm$estimates[which(qmsm$estimates>0)]), NA)
for (i in 1:length(vartime)){
trans[i,]= qmsm$estimates[which(qmsm$estimates>0)]
}
trans=as.data.frame(trans)
for (i in 1:length(qmsm$estimates[which(qmsm$estimates>0)]) ){
colnames(trans)[i]<-paste0("Haz_",tr_start_state[i],"_to_",tr_end_state[i])
}
}
if (ci.json=="none" ) {
trans=matrix(nrow=length(vartime), ncol=length(qmsm[which(qmsm>0)]), NA)
for (i in 1:length(vartime)){
trans[i,]= qmsm[which(qmsm>0)]
}
trans=as.data.frame(trans)
for (i in 1:length(qmsm[which(qmsm>0)]) ){
colnames(trans)[i]<-paste0("Haz_",tr_start_state[i],"_to_",tr_end_state[i])
}
}
trans$vartime=vartime
trans
##################### transitions lci and uci #####################################################################
if (ci.json!="none" ) {
trans_lci=matrix(nrow=length(vartime), ncol=length(qmsm$estimates[which(qmsm$L>0)]), NA)
for (i in 1:length(vartime)){
trans_lci[i,]= qmsm$L[which(qmsm$L>0)]
}
trans_lci=as.data.frame(trans_lci)
for (i in 1:length(qmsm$L[which(qmsm$L>0)]) ){
colnames(trans_lci)[i]<- paste0("Haz_",tr_start_state[i],"_to_",tr_end_state[i],"_lci")
}
trans_lci$vartime=vartime
trans_lci
trans_uci=matrix(nrow=length(vartime), ncol=length(qmsm$estimates[which(qmsm$L>0)]), NA)
for (i in 1:length(vartime)){
trans_uci[i,]= qmsm$L[which(qmsm$L>0)]
}
trans_uci=as.data.frame(trans_uci)
for (i in 1:length(qmsm$L[which(qmsm$L>0)]) ){
colnames(trans_uci)[i]<-paste0("Haz_",tr_start_state[i],"_to_",tr_end_state[i],"_uci")
}
trans_uci$vartime=vartime
trans_uci
}
#################################################################################################################
###Los ###
if (totlos==TRUE) {
los_array= array(dim=c(length(vartime),nstates+1,nstates),NA)
los_array_lci= array(dim=c(length(vartime),nstates+1,nstates),NA)
los_array_uci= array(dim=c(length(vartime),nstates+1,nstates),NA)
for (j in states) {
for (i in 1:length(vartime)) {
los.msm <- totlos.msm(msm.model, start=j, fromt=0, tot= vartime[i],ci=ci.json,cl=cl.json,
piecewise.times=piecewise.times.json, piecewise.covariates=piecewise.covariates.json,
num.integ=num.integ.json, B=B.json,cores=cores.json)
if (ci.json=="none" ) { los_array[i,,j]=cbind(t(los.msm),vartime=vartime[i] ) }
if (ci.json!="none" ) {
los_array[i,,j]=cbind(t(los.msm[1,]),vartime=vartime[i] )
los_array_lci[i,,j]=cbind(t(round(los.msm[2,],5)),vartime=vartime[i] )
los_array_uci[i,,j]=cbind(t(round(los.msm[3,],5)),vartime=vartime[i] )
}
}
}
loslist=list()
for (k in states) {
loslist[[k]]=los_array[,,k]
loslist[[k]]=as.data.frame(loslist[[k]])
}
if (length(vartime)==1) {
for (k in states) {
loslist[[k]]=as.data.frame(t(loslist[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(loslist[[k]])[j]=paste0("Los_",k,"_to_",j)
}
colnames(loslist[[k]])[nstates+1] ="vartime"
}
loslist
##################### los lci and uci #####################################################################
loslist_lci=list()
loslist_uci=list()
if (ci.json!="none" ) {
###lci
for (k in states) {
loslist_lci[[k]]=los_array_lci[,,k]
loslist_lci[[k]]=as.data.frame(loslist_lci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
loslist_lci[[k]]=as.data.frame(t(loslist_lci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(loslist_lci[[k]])[j]=paste0("Los_",k,"_to_",j,"_lci")
}
colnames(loslist_lci[[k]])[nstates+1] ="vartime"
}
loslist_lci
###uci
for (k in states) {
loslist_uci[[k]]=los_array_uci[,,k]
loslist_uci[[k]]=as.data.frame(loslist_uci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
loslist_uci[[k]]=as.data.frame(t(loslist_uci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(loslist_uci[[k]])[j]=paste0("Los_",k,"_to_",j,"_uci")
}
colnames(loslist_uci[[k]])[nstates+1] ="vartime"
}
loslist_uci
}
}
#################################################################################################################
###Sojourn ###
if (sojourn==TRUE) {
sojlist_=list()
sojlist_lci=list()
sojlist_uci=list()
soj.msm=vector()
if (ci.json!="none") {
soj.msm=sojourn.msm(msm.model,ci=ci.json,cl=cl.json, B=B.json)
sojnames= rownames(soj.msm)
sojlist=list()
sojlist[[1]]=cbind.data.frame(t( soj.msm$estimates))
names(sojlist[[1]])[1:(ncol(sojlist[[1]]))]<-paste0("Soj_",sub("State ","",sojnames))
sojlist
sojlist_lci[[1]]=cbind.data.frame(t(soj.msm$L))
sojlist_uci[[1]]=cbind.data.frame(t(soj.msm$U))
names(sojlist_lci[[1]])[1:(ncol(sojlist_lci[[1]]))]<-paste0("Soj_",sub("State ","",sojnames),"_lci")
names(sojlist_uci[[1]])[1:(ncol(sojlist_uci[[1]]))]<-paste0("Soj_",sub("State ","",sojnames),"_uci")
}
else if (ci.json =="none") {
soj.msm=sojourn.msm(msm.model,ci=ci.json,cl=cl.json, B=B.json)
sojlist=list()
sojlist[[1]]=cbind.data.frame(t( soj.msm$estimates))
sojnames=c(1:ncol(sojlist[[1]]))
names(sojlist[[1]])[1:(ncol(sojlist[[1]]))]<-paste0("Soj_",sub("State ","",sojnames))
sojlist
}
}
#################################################################################################################
###Next ###
########################### Main estimates ####################################################
if (pnext==TRUE) {
next_array= array(dim=c(1,nstates,nstates),NA)
next.msm <- pnext.msm(msm.model,ci=ci.json,cl=cl.json, B=B.json,cores=cores.json)
next_array[1,,]=t(cbind.data.frame(next.msm$estimates ))
nextlist=list()
for (k in states) {
nextlist[[k]]=next_array[,,k]
nextlist[[k]]=as.data.frame(t(nextlist[[k]]))
}
# if (length(vartime)==1) {
# for (k in states) {
# nextlist[[k]]=as.data.frame(t(nextlist[[k]]))
# }
# }
for (k in states) {
for (j in states) {
colnames(nextlist[[k]])[j]=paste0("Next_",k,"_to_",j)
}
}
nextlist
########################### next uci , lci ####################################################
nextlist_lci=list()
nextlist_uci=list()
if (ci.json!="none" ) {
###lci
next_array_lci= array(dim=c(1,nstates,nstates),NA)
next_array_lci[1,,]=t(cbind.data.frame(next.msm$L ))
for (k in states) {
nextlist_lci[[k]]=next_array_lci[,,k]
nextlist_lci[[k]]=as.data.frame(t(nextlist_lci[[k]]))
}
# if (length(vartime)==1) {
# for (k in states) {
# nextlist_lci[[k]]=as.data.frame(t(nextlist_lci[[k]]))
# }
# }
for (k in states) {
for (j in states) {
colnames(nextlist_lci[[k]])[j]=paste0("Next_",k,"_to_",j,"_lci")
}
}
nextlist_lci
### uci
next_array_uci= array(dim=c(1,nstates,nstates),NA)
next_array_uci[1,,]=t(cbind.data.frame(next.msm$U ))
for (k in states) {
nextlist_uci[[k]]=next_array_uci[,,k]
nextlist_uci[[k]]=as.data.frame(t(nextlist_uci[[k]]))
}
# if (length(vartime)==1) {
# for (k in states) {
# nextlist_uci[[k]]=as.data.frame(t(nextlist_uci[[k]]))
# }
# }
#
for (k in states) {
for (j in states) {
colnames(nextlist_uci[[k]])[j]=paste0("Next_",k,"_to_",j,"_uci")
}
}
nextlist_uci
}
}
###############################################################################
##First
if (efpt==TRUE) {
first.msm=vector()
first.msm= efpt.msm(msm.model, start="all", tostate=states,
ci=ci.json,cl=cl.json, B=B.json,cores=cores.json)
firstlist=list()
if (ci.json=="none" ) {
firstlist[[1]]=cbind.data.frame(t( first.msm))
}
if (ci.json!="none" ) {
firstlist[[1]]=cbind.data.frame(t( first.msm[1,]))
}
for (i in 1:nstates) {
names( firstlist[[1]])[i]<-paste0("First_",i)
}
firstlist_lci=list()
firstlist_uci=list()
if (ci.json!="none" ) {
firstlist_lci[[1]]=cbind.data.frame(t(first.msm[2,]))
firstlist_uci[[1]]=cbind.data.frame(t(first.msm[3,]))
for (i in 1:nstates) {
names( firstlist_lci[[1]])[i]<-paste0("First_",i,"_lci")
names( firstlist_uci[[1]])[i]<-paste0("First_",i,"_uci")
}
}
}
##########################################################################################
##Number
if (envisits==TRUE) {
############Main estimates ########################
number_array= array(dim=c(length(vartime),nstates+1,nstates),NA)
number_array_lci= array(dim=c(length(vartime),nstates+1,nstates),NA)
number_array_uci= array(dim=c(length(vartime),nstates+1,nstates),NA)
for (j in states) {
for (i in 1:length(vartime)) {
number.msm <- envisits.msm(msm.model, start=j, fromt=0, tot= vartime[i],
piecewise.times=piecewise.times.json, piecewise.covariates=piecewise.covariates.json,
num.integ=num.integ.json,
ci=ci.json, cl=cl.json, B=B.json,
cores=cores.json)
if (ci.json=="none" ) { number_array[i,,j]=cbind(t(number.msm),vartime=vartime[i] ) }
if (ci.json!="none" ) {
number_array[i,,j]=cbind(t(round(number.msm[1,],5)),vartime=vartime[i] )
number_array_lci[i,,j]=cbind(t(round(number.msm[2,],5)),vartime=vartime[i] )
number_array_uci[i,,j]=cbind(t(round(number.msm[3,],5)),vartime=vartime[i] )
}
}
}
numberlist=list()
for (k in states) {
numberlist[[k]]=number_array[,,k]
numberlist[[k]]=as.data.frame(numberlist[[k]])
}
if (length(vartime)==1) {
for (k in states) {
numberlist[[k]]=as.data.frame(t(numberlist[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(numberlist[[k]])[j]=paste0("Number_",k,"_to_",j)
}
colnames(numberlist[[k]])[nstates+1] ="vartime"
}
numberlist
numberlist_lci=list()
numberlist_uci=list()
if (ci.json!="none" ) {
############ lci estimates ########################
for (k in states) {
numberlist_lci[[k]]=number_array_lci[,,k]
numberlist_lci[[k]]=as.data.frame(numberlist_lci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
numberlist_lci[[k]]=as.data.frame(t(numberlist_lci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(numberlist_lci[[k]])[j]=paste0("Number_",k,"_to_",j,"_lci")
}
colnames(numberlist_lci[[k]])[nstates+1] ="vartime"
}
numberlist_lci
############ uci estimates ########################
for (k in states) {
numberlist_uci[[k]]=number_array_uci[,,k]
numberlist_uci[[k]]=as.data.frame(numberlist_uci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
numberlist_uci[[k]]=as.data.frame(t(numberlist_uci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(numberlist_uci[[k]])[j]=paste0("Number_",k,"_to_",j,"_uci")
}
colnames(numberlist_uci[[k]])[nstates+1] ="vartime"
}
numberlist_uci
}
}
###########################################################################################################
##Visit
if (visit==TRUE) {
visit.msm=vector()
######## What should we put in start? 1 or all states
visit_array= array(dim=c(length(vartime),nstates+1,nstates),NA)
visit_array_lci= array(dim=c(length(vartime),nstates+1,nstates),NA)
visit_array_uci= array(dim=c(length(vartime),nstates+1,nstates),NA)
for (i in 1:length(vartime)) {
visit.msm <- ppass.msm(x=msm.model, qmatrix=NULL, tot=vartime[i], start="all",
piecewise.times=piecewise.times.json, piecewise.covariates=piecewise.covariates.json,
ci=ci.json, cl=cl.json, B=B.json,
cores=cores.json)
if (ci.json=="none" ) {visit_array[i,,]=t(cbind(visit.msm,vartime=vartime[i] )) }
if (ci.json!="none" ) {
visit_array[i,,]=t(cbind(visit.msm$estimates,vartime=vartime[i] ))
visit_array_lci[i,,]=t(cbind(visit.msm$L,vartime=vartime[i] ))
visit_array_uci[i,,]=t(cbind(visit.msm$U,vartime=vartime[i] ))
}
}
####################### Main estimates ########################
vlist=list()
for (k in states) {
vlist[[k]]=visit_array[,,k]
vlist[[k]]=as.data.frame(vlist[[k]])
}
if (length(vartime)==1) {
for (k in states) {
vlist[[k]]=as.data.frame(t(vlist[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(vlist[[k]])[j]=paste0("Visit_",k,"_to_",j)
}
colnames(vlist[[k]])[nstates+1] ="vartime"
}
vlist
vlist_lci=list()
vlist_uci=list()
if (ci.json!="none" ) {
####################### Lci estimates ########################
for (k in states) {
vlist_lci[[k]]=visit_array_lci[,,k]
vlist_lci[[k]]=as.data.frame(vlist_lci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
vlist_lci[[k]]=as.data.frame(t(vlist_lci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(vlist_lci[[k]])[j]=paste0("Visit_",k,"_to_",j,"_lci")
}
colnames(vlist_lci[[k]])[nstates+1] ="vartime"
}
vlist_lci
####################### Uci estimates ########################
for (k in states) {
vlist_uci[[k]]=visit_array_uci[,,k]
vlist_uci[[k]]=as.data.frame(vlist_uci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
vlist_uci[[k]]=as.data.frame(t(vlist_uci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(vlist_uci[[k]])[j]=paste0("Visit_",k,"_to_",j,"_uci")
}
colnames(vlist_uci[[k]])[nstates+1] ="vartime"
}
vlist_uci
}
}
###########################################################################################
# ###Ratio of tr. intensities
# mat
#
# p=1
# qratio=matrix(nrow=1, ncol=length(which(mat>0))^2,NA)
# names.qratio=vector()
#
# for (i in 1:nstates) {
# for (j in 1:nstates) {
# for (k in 1:nstates) {
# for (f in 1:nstates) {
#
# if (mat[k,f]==0 | mat[i,j]==0 | k==f | i==j) {next}
# else if (mat[k,f]!=0 & mat[i,j]!=0 & k!=f & i!=j ) {
# qratio[1,p]=qratio.msm(msm.model, ind1=c(i,j), ind2=c(k,f), covariates = covariates_list[[g]])[1]
# names.qratio[p]=paste0("Qratio_t1_",i,".",j,"_t2_",k,".",f)
# p=p+1
# }
# }
# }
# }
# }
#
# qratio=as.data.frame(qratio)
# names(qratio)=names.qratio
#
# qratiolist=list()
# qratiolist[[1]]=cbind.data.frame(qratio,cov_factor=rep(paste0(names(covariates_list[[g]]),covariates_list[[g]]),1))
##############################################################################################
pred[[1]]= list(probs=plist,probs_lci=plist_lci,probs_uci=plist_uci,
trans=trans, trans_lci=trans_lci,trans_uci=trans_uci,
los=loslist,los_lci=loslist_lci,los_uci=loslist_uci,
soj=sojlist,soj_lci=sojlist_lci,soj_uci=sojlist_uci,
nextv=nextlist,nextv_lci=nextlist_lci,nextv_uci=nextlist_uci,
first=firstlist,first_lci=firstlist_lci,first_uci=firstlist_uci,
number=numberlist,number_lci=numberlist_lci,number_uci=numberlist_uci,
visit=vlist,visit_lci=vlist_lci, visit_uci=vlist_uci
)
}
if (length(covariates_list) != 0) {
for (g in 1:length(covariates_list)) {
#g=1
################################################################################################################
### Create all the lists to be output even if they will remain vacant ###
loslist=list()
loslist_lci=list()
loslist_uci=list()
sojlist=list()
sojlist_lci=list()
sojlist_uci=list()
nextlist=list()
nextlist_lci=list()
nextlist_uci=list()
firstlist=list()
firstlist_lci=list()
firstlist_uci=list()
numberlist=list()
numberlist_lci=list()
numberlist_uci=list()
vlist=list()
vlist_lci=list()
vlist_uci=list()
trans_lci=vector()
trans_uci=vector()
#################################################################################################################
###Probability ###
mat=mat.init
mat[is.na(mat)] <- 0
ntransitions=length(mat[which(mat>0)])
nstates=ncol(mat)
states=c(seq(1:nstates))
##################### Main estimates#####################################################################
p_array= array(dim=c(length(vartime),nstates+1,nstates),NA)
p_array_lci= array(dim=c(length(vartime),nstates+1,nstates),NA)
p_array_uci= array(dim=c(length(vartime),nstates+1,nstates),NA)
for (i in 1:length(vartime)) {
p.msm <- pmatrix.msm(msm.model,t=vartime[i], t1=0, ci=ci.json , cl=cl.json, B=B.json,
cores=cores.json ,covariates= covariates_list[[g]])
if (ci.json=="none" ) {
p_array[i,,]=t(cbind(p.msm,vartime=vartime[i] ))
}
if (ci.json!="none" ) {
p_array[i,,]=t(cbind(p.msm$estimates,vartime=vartime[i] ))
p_array_lci[i,,]=t(cbind(p.msm$L,vartime=vartime[i] ))
p_array_uci[i,,]=t(cbind(p.msm$U,vartime=vartime[i] ))
}
}
plist=list()
for (k in states) {
plist[[k]]=p_array[,,k]
plist[[k]]=as.data.frame(plist[[k]])
}
if (length(vartime)==1) {
for (k in states) {
plist[[k]]=as.data.frame(t(plist[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(plist[[k]])[j]=paste0("P_",k,"_to_",j)
}
colnames(plist[[k]])[nstates+1] ="vartime"
}
plist
plist_lci=list()
plist_uci=list()
if (ci.json!="none" ) {
##################### Probabilities lci #####################################################################
for (k in states) {
plist_lci[[k]]=p_array_lci[,,k]
plist_lci[[k]]=as.data.frame(plist_lci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
plist_lci[[k]]=as.data.frame(t(plist_lci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(plist_lci[[k]])[j]=paste0("P_",k,"_to_",j,"_lci")
}
colnames(plist_lci[[k]])[nstates+1] ="vartime"
}
plist_lci
##################### Probabilities uci #####################################################################
plist_uci=list()
for (k in states) {
plist_uci[[k]]=p_array_uci[,,k]
plist_uci[[k]]=as.data.frame(plist_uci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
plist_uci[[k]]=as.data.frame(t(plist_uci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(plist_uci[[k]])[j]=paste0("P_",k,"_to_",j,"_uci")
}
colnames(plist_uci[[k]])[nstates+1] ="vartime"
}
plist_uci
}
#################################################################################################################
###Transitions values for the period we want (of course they are constant)###
##################### Main estimates#####################################################################
qmsm=qmatrix.msm(msm.model,covariates=covariates_list[[g]],ci=ci.json , cl=cl.json,
B=B.json, cores=cores.json)
if (ci.json!="none" ) {
trans=matrix(nrow=length(vartime), ncol=length(qmsm$estimates[which(qmsm$estimates>0)]), NA)
for (i in 1:length(vartime)){
trans[i,]= qmsm$estimates[which(qmsm$estimates>0)]
}
trans=as.data.frame(trans)
for (i in 1:length(qmsm$estimates[which(qmsm$estimates>0)]) ){
colnames(trans)[i]<-paste0("Haz_",tr_start_state[i],"_to_",tr_end_state[i])
}
}
if (ci.json=="none" ) {
trans=matrix(nrow=length(vartime), ncol=length(qmsm[which(qmsm>0)]), NA)
for (i in 1:length(vartime)){
trans[i,]= qmsm[which(qmsm>0)]
}
trans=as.data.frame(trans)
for (i in 1:length(qmsm[which(qmsm>0)]) ){
colnames(trans)[i]<-paste0("Haz_",tr_start_state[i],"_to_",tr_end_state[i])
}
}
trans$vartime=vartime
trans
##################### transitions lci and uci #####################################################################
if (ci.json!="none" ) {
trans_lci=matrix(nrow=length(vartime), ncol=length(qmsm$estimates[which(qmsm$L>0)]), NA)
for (i in 1:length(vartime)){
trans_lci[i,]= qmsm$L[which(qmsm$L>0)]
}
trans_lci=as.data.frame(trans_lci)
for (i in 1:length(qmsm$L[which(qmsm$L>0)]) ){
colnames(trans_lci)[i]<- paste0("Haz_",tr_start_state[i],"_to_",tr_end_state[i],"_lci")
}
trans_lci$vartime=vartime
trans_lci
trans_uci=matrix(nrow=length(vartime), ncol=length(qmsm$estimates[which(qmsm$L>0)]), NA)
for (i in 1:length(vartime)){
trans_uci[i,]= qmsm$L[which(qmsm$L>0)]
}
trans_uci=as.data.frame(trans_uci)
for (i in 1:length(qmsm$L[which(qmsm$L>0)]) ){
colnames(trans_uci)[i]<-paste0("Haz_",tr_start_state[i],"_to_",tr_end_state[i],"_uci")
}
trans_uci$vartime=vartime
trans_uci
}
#################################################################################################################
###Los ###
if (totlos==TRUE) {
los_array= array(dim=c(length(vartime),nstates+1,nstates),NA)
los_array_lci= array(dim=c(length(vartime),nstates+1,nstates),NA)
los_array_uci= array(dim=c(length(vartime),nstates+1,nstates),NA)
for (j in states) {
for (i in 1:length(vartime)) {
los.msm <- totlos.msm(msm.model, start=j, fromt=0, tot= vartime[i],covariates=covariates_list[[g]],ci=ci.json,cl=cl.json,
piecewise.times=piecewise.times.json, piecewise.covariates=piecewise.covariates.json,
num.integ=num.integ.json, B=B.json,cores=cores.json)
if (ci.json=="none" ) { los_array[i,,j]=cbind(t(los.msm),vartime=vartime[i] ) }
if (ci.json!="none" ) {
los_array[i,,j]=cbind(t(los.msm[1,]),vartime=vartime[i] )
los_array_lci[i,,j]=cbind(t(round(los.msm[2,],5)),vartime=vartime[i] )
los_array_uci[i,,j]=cbind(t(round(los.msm[3,],5)),vartime=vartime[i] )
}
}
}
loslist=list()
for (k in states) {
loslist[[k]]=los_array[,,k]
loslist[[k]]=as.data.frame(loslist[[k]])
}
if (length(vartime)==1) {
for (k in states) {
loslist[[k]]=as.data.frame(t(loslist[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(loslist[[k]])[j]=paste0("Los_",k,"_to_",j)
}
colnames(loslist[[k]])[nstates+1] ="vartime"
}
loslist
##################### los lci and uci #####################################################################
loslist_lci=list()
loslist_uci=list()
if (ci.json!="none" ) {
###lci
for (k in states) {
loslist_lci[[k]]=los_array_lci[,,k]
loslist_lci[[k]]=as.data.frame(loslist_lci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
loslist_lci[[k]]=as.data.frame(t(loslist_lci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(loslist_lci[[k]])[j]=paste0("Los_",k,"_to_",j,"_lci")
}
colnames(loslist_lci[[k]])[nstates+1] ="vartime"
}
loslist_lci
###uci
for (k in states) {
loslist_uci[[k]]=los_array_uci[,,k]
loslist_uci[[k]]=as.data.frame(loslist_uci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
loslist_uci[[k]]=as.data.frame(t(loslist_uci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(loslist_uci[[k]])[j]=paste0("Los_",k,"_to_",j,"_uci")
}
colnames(loslist_uci[[k]])[nstates+1] ="vartime"
}
loslist_uci
}
}
#################################################################################################################
###Sojourn ###
if (sojourn==TRUE) {
sojlist_=list()
sojlist_lci=list()
sojlist_uci=list()
soj.msm=vector()
if (ci.json!="none") {
soj.msm=sojourn.msm(msm.model,covariates=covariates_list[[g]],ci=ci.json,cl=cl.json, B=B.json)
sojnames= rownames(soj.msm)
sojlist=list()
sojlist[[1]]=cbind.data.frame(t( soj.msm$estimates))
names(sojlist[[1]])[1:(ncol(sojlist[[1]]))]<-paste0("Soj_",sub("State ","",sojnames))
sojlist
sojlist_lci[[1]]=cbind.data.frame(t(soj.msm$L))
sojlist_uci[[1]]=cbind.data.frame(t(soj.msm$U))
names(sojlist_lci[[1]])[1:(ncol(sojlist_lci[[1]]))]<-paste0("Soj_",sub("State ","",sojnames),"_lci")
names(sojlist_uci[[1]])[1:(ncol(sojlist_uci[[1]]))]<-paste0("Soj_",sub("State ","",sojnames),"_uci")
}
else if (ci.json =="none") {
soj.msm=sojourn.msm(msm.model,covariates=covariates_list[[g]],ci=ci.json,cl=cl.json, B=B.json)
sojlist=list()
sojlist[[1]]=cbind.data.frame(t( soj.msm$estimates))
sojnames=c(1:ncol(sojlist[[1]]))
names(sojlist[[1]])[1:(ncol(sojlist[[1]]))]<-paste0("Soj_",sub("State ","",sojnames))
sojlist
}
}
#################################################################################################################
###Next ###
########################### Main estimates ####################################################
if (pnext==TRUE) {
next_array= array(dim=c(1,nstates,nstates),NA)
next.msm <- pnext.msm(msm.model,ci=ci.json,cl=cl.json, B=B.json,cores=cores.json, covariates=covariates_list[[g]])
next_array[1,,]=t(cbind.data.frame(next.msm$estimates ))
nextlist=list()
for (k in states) {
nextlist[[k]]=next_array[,,k]
nextlist[[k]]=as.data.frame(t(nextlist[[k]]))
}
# if (length(vartime)==1) {
# for (k in states) {
# nextlist[[k]]=as.data.frame(t(nextlist[[k]]))
# }
# }
for (k in states) {
for (j in states) {
colnames(nextlist[[k]])[j]=paste0("Next_",k,"_to_",j)
}
}
nextlist
########################### next uci , lci ####################################################
nextlist_lci=list()
nextlist_uci=list()
if (ci.json!="none" ) {
###lci
next_array_lci= array(dim=c(1,nstates,nstates),NA)
next_array_lci[1,,]=t(cbind.data.frame(next.msm$L ))
for (k in states) {
nextlist_lci[[k]]=next_array_lci[,,k]
nextlist_lci[[k]]=as.data.frame(t(nextlist_lci[[k]]))
}
# if (length(vartime)==1) {
# for (k in states) {
# nextlist_lci[[k]]=as.data.frame(t(nextlist_lci[[k]]))
# }
# }
for (k in states) {
for (j in states) {
colnames(nextlist_lci[[k]])[j]=paste0("Next_",k,"_to_",j,"_lci")
}
}
nextlist_lci
### uci
next_array_uci= array(dim=c(1,nstates,nstates),NA)
next_array_uci[1,,]=t(cbind.data.frame(next.msm$U ))
for (k in states) {
nextlist_uci[[k]]=next_array_uci[,,k]
nextlist_uci[[k]]=as.data.frame(t(nextlist_uci[[k]]))
}
# if (length(vartime)==1) {
# for (k in states) {
# nextlist_uci[[k]]=as.data.frame(t(nextlist_uci[[k]]))
# }
# }
for (k in states) {
for (j in states) {
colnames(nextlist_uci[[k]])[j]=paste0("Next_",k,"_to_",j,"_uci")
}
}
nextlist_uci
}
}
###############################################################################
##First
if (efpt==TRUE) {
first.msm=vector()
first.msm= efpt.msm(msm.model, start="all", tostate=states,
ci=ci.json,cl=cl.json, B=B.json,cores=cores.json, covariates = covariates_list[[g]])
firstlist=list()
if (ci.json=="none" ) {
firstlist[[1]]=cbind.data.frame(t( first.msm))
}
if (ci.json!="none" ) {
firstlist[[1]]=cbind.data.frame(t( first.msm[1,]))
}
for (i in 1:nstates) {
names( firstlist[[1]])[i]<-paste0("First_",i)
}
firstlist_lci=list()
firstlist_uci=list()
if (ci.json!="none" ) {
firstlist_lci[[1]]=cbind.data.frame(t(first.msm[2,]))
firstlist_uci[[1]]=cbind.data.frame(t(first.msm[3,]))
for (i in 1:nstates) {
names( firstlist_lci[[1]])[i]<-paste0("First_",i,"_lci")
names( firstlist_uci[[1]])[i]<-paste0("First_",i,"_uci")
}
}
}
##########################################################################################
##Number
if (envisits==TRUE) {
############Main estimates ########################
number_array= array(dim=c(length(vartime),nstates+1,nstates),NA)
number_array_lci= array(dim=c(length(vartime),nstates+1,nstates),NA)
number_array_uci= array(dim=c(length(vartime),nstates+1,nstates),NA)
for (j in states) {
for (i in 1:length(vartime)) {
number.msm <- envisits.msm(msm.model, start=j, fromt=0, tot= vartime[i],covariates=covariates_list[[g]],
piecewise.times=piecewise.times.json, piecewise.covariates=piecewise.covariates.json,
num.integ=num.integ.json,
ci=ci.json, cl=cl.json, B=B.json,
cores=cores.json)
if (ci.json=="none" ) { number_array[i,,j]=cbind(t(number.msm),vartime=vartime[i] ) }
if (ci.json!="none" ) {
number_array[i,,j]=cbind(t(round(number.msm[1,],5)),vartime=vartime[i] )
number_array_lci[i,,j]=cbind(t(round(number.msm[2,],5)),vartime=vartime[i] )
number_array_uci[i,,j]=cbind(t(round(number.msm[3,],5)),vartime=vartime[i] )
}
}
}
numberlist=list()
for (k in states) {
numberlist[[k]]=number_array[,,k]
numberlist[[k]]=as.data.frame(numberlist[[k]])
}
if (length(vartime)==1) {
for (k in states) {
numberlist[[k]]=as.data.frame(t(numberlist[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(numberlist[[k]])[j]=paste0("Number_",k,"_to_",j)
}
colnames(numberlist[[k]])[nstates+1] ="vartime"
}
numberlist
numberlist_lci=list()
numberlist_uci=list()
if (ci.json!="none" ) {
############ lci estimates ########################
for (k in states) {
numberlist_lci[[k]]=number_array_lci[,,k]
numberlist_lci[[k]]=as.data.frame(numberlist_lci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
numberlist_lci[[k]]=as.data.frame(t(numberlist_lci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(numberlist_lci[[k]])[j]=paste0("Number_",k,"_to_",j,"_lci")
}
colnames(numberlist_lci[[k]])[nstates+1] ="vartime"
}
numberlist_lci
############ uci estimates ########################
for (k in states) {
numberlist_uci[[k]]=number_array_uci[,,k]
numberlist_uci[[k]]=as.data.frame(numberlist_uci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
numberlist_uci[[k]]=as.data.frame(t(numberlist_uci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(numberlist_uci[[k]])[j]=paste0("Number_",k,"_to_",j,"_uci")
}
colnames(numberlist_uci[[k]])[nstates+1] ="vartime"
}
numberlist_uci
}
}
###########################################################################################################
##Visit
if (visit==TRUE) {
visit.msm=vector()
######## What should we put in start? 1 or all states
visit_array= array(dim=c(length(vartime),nstates+1,nstates),NA)
visit_array_lci= array(dim=c(length(vartime),nstates+1,nstates),NA)
visit_array_uci= array(dim=c(length(vartime),nstates+1,nstates),NA)
for (i in 1:length(vartime)) {
visit.msm <- ppass.msm(x=msm.model, qmatrix=NULL, tot=vartime[i], start="all", covariates = covariates_list[[g]],
piecewise.times=piecewise.times.json, piecewise.covariates=piecewise.covariates.json,
ci=ci.json, cl=cl.json, B=B.json,
cores=cores.json)
if (ci.json=="none" ) {visit_array[i,,]=t(cbind(visit.msm,vartime=vartime[i] )) }
if (ci.json!="none" ) {
visit_array[i,,]=t(cbind(visit.msm$estimates,vartime=vartime[i] ))
visit_array_lci[i,,]=t(cbind(visit.msm$L,vartime=vartime[i] ))
visit_array_uci[i,,]=t(cbind(visit.msm$U,vartime=vartime[i] ))
}
}
####################### Main estimates ########################
vlist=list()
for (k in states) {
vlist[[k]]=visit_array[,,k]
vlist[[k]]=as.data.frame(vlist[[k]])
}
if (length(vartime)==1) {
for (k in states) {
vlist[[k]]=as.data.frame(t(vlist[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(vlist[[k]])[j]=paste0("Visit_",k,"_to_",j)
}
colnames(vlist[[k]])[nstates+1] ="vartime"
}
vlist
vlist_lci=list()
vlist_uci=list()
if (ci.json!="none" ) {
####################### Lci estimates ########################
for (k in states) {
vlist_lci[[k]]=visit_array_lci[,,k]
vlist_lci[[k]]=as.data.frame(vlist_lci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
vlist_lci[[k]]=as.data.frame(t(vlist_lci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(vlist_lci[[k]])[j]=paste0("Visit_",k,"_to_",j,"_lci")
}
colnames(vlist_lci[[k]])[nstates+1] ="vartime"
}
vlist_lci
####################### Uci estimates ########################
for (k in states) {
vlist_uci[[k]]=visit_array_uci[,,k]
vlist_uci[[k]]=as.data.frame(vlist_uci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
vlist_uci[[k]]=as.data.frame(t(vlist_uci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(vlist_uci[[k]])[j]=paste0("Visit_",k,"_to_",j,"_uci")
}
colnames(vlist_uci[[k]])[nstates+1] ="vartime"
}
vlist_uci
}
}
###########################################################################################
# ###Ratio of tr. intensities
# mat
#
# p=1
# qratio=matrix(nrow=1, ncol=length(which(mat>0))^2,NA)
# names.qratio=vector()
#
# for (i in 1:nstates) {
# for (j in 1:nstates) {
# for (k in 1:nstates) {
# for (f in 1:nstates) {
#
# if (mat[k,f]==0 | mat[i,j]==0 | k==f | i==j) {next}
# else if (mat[k,f]!=0 & mat[i,j]!=0 & k!=f & i!=j ) {
# qratio[1,p]=qratio.msm(msm.model, ind1=c(i,j), ind2=c(k,f), covariates = covariates_list[[g]])[1]
# names.qratio[p]=paste0("Qratio_t1_",i,".",j,"_t2_",k,".",f)
# p=p+1
# }
# }
# }
# }
# }
#
# qratio=as.data.frame(qratio)
# names(qratio)=names.qratio
#
# qratiolist=list()
# qratiolist[[1]]=cbind.data.frame(qratio,cov_factor=rep(paste0(names(covariates_list[[g]]),covariates_list[[g]]),1))
##############################################################################################
pred[[g]]= list(probs=plist,probs_lci=plist_lci,probs_uci=plist_uci,
trans=trans, trans_lci=trans_lci,trans_uci=trans_uci,
los=loslist,los_lci=loslist_lci,los_uci=loslist_uci,
soj=sojlist,soj_lci=sojlist_lci,soj_uci=sojlist_uci,
nextv=nextlist,nextv_lci=nextlist_lci,nextv_uci=nextlist_uci,
first=firstlist,first_lci=firstlist_lci,first_uci=firstlist_uci,
number=numberlist,number_lci=numberlist_lci,number_uci=numberlist_uci,
visit=vlist,visit_lci=vlist_lci, visit_uci=vlist_uci
)
}
}
########################################################################################################
######## Declare the list elements even if they stay empty ####
pjson_lci=list()
pjson_uci=list()
hjson_lci=list()
hjson_uci=list()
losjson=list()
losjson_lci=list()
losjson_uci=list()
sojjson=list()
sojjson_lci=list()
sojjson_uci=list()
nextjson=list()
nextjson_lci=list()
nextjson_uci=list()
firstjson=list()
firstjson_lci=list()
firstjson_uci=list()
numberjson=list()
numberjson_lci=list()
numberjson_uci=list()
vjson=list()
vjson_lci=list()
vjson_uci=list()
########################################################################################################
#### Probabilities
########################################################################################################
##### Main estimates
pjson=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
pjson[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$probs[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$probs[[k]])[i]), pjson[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplist=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplist[1,]=as.vector(pred[[1]]$probs[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplist[j,]=as.vector(pred[[j]]$probs[[k]][,i])
}
pjson[[p]]=tmplist
assign(paste0("forMSM",names(pred[[1]]$probs[[k]])[i]), pjson[[p]])
p=p+1
}
}
}
pvector=vector()
pvector=ls(pattern = "forMSMP_._to_.$")
pvector=sub("forMSM","",ls(pattern = "^forMSMP_._to_.$") )
pvector=stringi::stri_sort(pvector, numeric = TRUE)
names(pjson)=pvector[1:(p-1)]
pjson
##### Lci possibilities estimates
if (ci.json!="none" ) {
pjson_lci=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
pjson_lci[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$probs_lci[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$probs_lci[[k]])[i]), pjson_lci[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplist=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplist[1,]=as.vector(pred[[1]]$probs_lci[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplist[j,]=as.vector(pred[[j]]$probs_lci[[k]][,i])
}
pjson_lci[[p]]=tmplist
assign(paste0("forMSM",names(pred[[1]]$probs_lci[[k]])[i]), pjson_lci[[p]])
p=p+1
}
}
}
pvector_lci=vector()
pvector_lci=ls(pattern = "^forMSMP.*lci$")
pvector_lci=sub("forMSM","",ls(pattern = "^forMSMP.*lci$") )
pvector_lci=stringi::stri_sort(pvector_lci, numeric = TRUE)
names(pjson_lci)=pvector_lci[1:(p-1)]
##### Uci possibilities estimates
pjson_uci=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
pjson_uci[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$probs_uci[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$probs_uci[[k]])[i]), pjson_uci[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplist=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplist[1,]=as.vector(pred[[1]]$probs_uci[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplist[j,]=as.vector(pred[[j]]$probs_uci[[k]][,i])
}
pjson_uci[[p]]=tmplist
assign(paste0("forMSM",names(pred[[1]]$probs_uci[[k]])[i]), pjson_uci[[p]])
p=p+1
}
}
}
pvector_uci=vector()
pvector_uci=ls(pattern = "^forMSMP.*uci$")
pvector_uci=sub("forMSM","",ls(pattern = "^forMSMP.*uci$") )
pvector_uci=stringi::stri_sort(pvector_uci, numeric = TRUE)
names(pjson_uci)=pvector_uci[1:(p-1)]
}
#################################################################################################################
### Transitions main estimates ###
hjson=list()
if (length(covariates_list)<=1) {
for (k in 1:ntransitions) {
tmplisth=matrix(nrow =1 , ncol= length(vartime))
tmplisth[1,]=as.vector(pred[[1]]$trans[,k])
hjson[[1]]=tmplisth
assign(paste0("forMSM",names(pred[[1]]$trans)[i]), hjson[[1]])
}
}
if (length(covariates_list)>1) {
for (k in 1:ntransitions) {
tmplisth=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplisth[1,]=as.vector(pred[[1]]$trans[,k])
for (j in 2:length(covariates_list)) {
tmplisth[j,]=as.vector(pred[[j]]$trans[,k])
}
hjson[[k]]=tmplisth
assign(paste0("forMSM",names(pred[[1]]$trans)[k]), hjson[[k]])
}
}
hvector=vector()
hvector=ls(pattern = "^forMSMHaz_._to_.$")
hvector=sub("forMSM","",ls(pattern = "^forMSMHaz_._to_.$") )
hvector=stringi::stri_sort(hvector, numeric = TRUE)
names(hjson)=hvector
if (ci.json!="none" ) {
### Transitions lci estimates ###
hjson_lci=list()
if (length(covariates_list)<=1) {
for (k in 1:ntransitions) {
tmplisth=matrix(nrow =1 , ncol= length(vartime))
tmplisth[1,]=as.vector(pred[[1]]$trans_lci[,k])
hjson_lci[[1]]=tmplisth
assign(paste0("forMSM",names(pred[[1]]$trans_lci)[i]), hjson_lci[[1]])
}
}
if (length(covariates_list)>1) {
for (k in 1:ntransitions) {
tmplisth=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplisth[1,]=as.vector(pred[[1]]$trans_lci[,k])
for (j in 2:length(covariates_list)) {
tmplisth[j,]=as.vector(pred[[j]]$trans_lci[,k])
}
hjson_lci[[k]]=tmplisth
assign(paste0("forMSM",names(pred[[1]]$trans_lci)[k]), hjson_lci[[k]])
}
}
hvector_lci=vector()
hvector_lci=ls(pattern = "^forMSMHaz_._to_._lci$")
hvector_lci=sub("forMSM","",ls(pattern = "^forMSMHaz_._to_._lci$") )
hvector_lci=stringi::stri_sort(hvector_lci, numeric = TRUE)
names(hjson_lci)=hvector_lci
### Transitions uci estimates ###
hjson_uci=list()
if (length(covariates_list)<=1) {
for (k in 1:ntransitions) {
tmplisth=matrix(nrow =1 , ncol= length(vartime))
tmplisth[1,]=as.vector(pred[[1]]$trans_uci[,k])
hjson_uci[[1]]=tmplisth
assign(paste0("forMSM",names(pred[[1]]$trans_uci)[i]), hjson_uci[[1]])
}
}
if (length(covariates_list)>1) {
for (k in 1:ntransitions) {
tmplisth=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplisth[1,]=as.vector(pred[[1]]$trans_uci[,k])
for (j in 2:length(covariates_list)) {
tmplisth[j,]=as.vector(pred[[j]]$trans_uci[,k])
}
hjson_uci[[k]]=tmplisth
assign(paste0("forMSM",names(pred[[1]]$trans_uci)[k]), hjson_uci[[k]])
}
}
hvector_uci=vector()
hvector_uci=ls(pattern = "^forMSMHaz_._to_._uci$")
hvector_uci=sub("forMSM","",ls(pattern = "^forMSMHaz_._to_._uci$") )
hvector_uci=stringi::stri_sort(hvector_uci, numeric = TRUE)
names(hjson_uci)=hvector_uci
}
##########################################################################
### Los ################################################################
##############################################################################
if (totlos==TRUE) {
### Los Main estimates #####
losjson=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
losjson[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$los[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$los[[k]])[i]), losjson[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplistlos=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplistlos[1,]=as.vector(pred[[1]]$los[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplistlos[j,]=as.vector(pred[[j]]$los[[k]][,i])
}
losjson[[p]]=tmplistlos
assign(paste0("forMSM",names(pred[[1]]$los[[k]])[i]), losjson[[p]])
p=p+1
}
}
}
losvector=vector()
losvector=ls(pattern = "forMSMLos_._to_.$")
losvector=sub("forMSM","",ls(pattern = "^forMSMLos_._to_.$") )
losvector=stringi::stri_sort(losvector, numeric = TRUE)
names(losjson)=losvector
if (ci.json!="none" ) {
### Los lci estimates #####
losjson_lci=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
losjson_lci[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$los_lci[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$los_lci[[k]])[i]), losjson_lci[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplistlos=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplistlos[1,]=as.vector(pred[[1]]$los_lci[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplistlos[j,]=as.vector(pred[[j]]$los_lci[[k]][,i])
}
losjson_lci[[p]]=tmplistlos
assign(paste0("forMSM",names(pred[[1]]$los_lci[[k]])[i]), losjson_lci[[p]])
p=p+1
}
}
}
losvector_lci=vector()
losvector_lci=ls(pattern = "^forMSMLos.*lci$")
losvector_lci=sub("forMSM","",ls(pattern = "^forMSMLos.*lci$") )
losvector_lci=stringi::stri_sort(losvector_lci, numeric = TRUE)
names(losjson_lci)=losvector_lci
### Los uci estimates #####
losjson_uci=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
losjson_uci[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$los_uci[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$los_uci[[k]])[i]), losjson_uci[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplistlos=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplistlos[1,]=as.vector(pred[[1]]$los_uci[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplistlos[j,]=as.vector(pred[[j]]$los_uci[[k]][,i])
}
losjson_uci[[p]]=tmplistlos
assign(paste0("forMSM",names(pred[[1]]$los_uci[[k]])[i]), losjson_uci[[p]])
p=p+1
}
}
}
losvector_uci=vector()
losvector_uci=ls(pattern = "^forMSMLos.*uci$")
losvector_uci=sub("forMSM","",ls(pattern = "^forMSMLos.*uci$") )
losvector_uci=stringi::stri_sort(losvector_uci, numeric = TRUE)
names(losjson_uci)=losvector_uci
}
}
##########################################################################
### Sojourn ################################################################
##############################################################################
if ( sojourn==TRUE) {
### Sojourn Main estimates #####
sojjson=list()
if (length(covariates_list)<=1) {
p=1
for (k in na_states) {
tmplistsoj=matrix(nrow =1 , ncol= 1)
tmplistsoj[1,1]=as.matrix(pred[[1]]$soj[[1]][1])
sojjson[[1]]=tmplistsoj
assign(paste0("forMSM",names(pred[[1]]$soj)[i]), tmplistsoj[[1]])
p=p+1
}
}
if (length(covariates_list)>1) {
p=1
for (k in na_states) {
tmplistsoj=matrix(nrow =length(covariates_list) , ncol= 1)
for (j in 1:length(covariates_list)) {
tmplistsoj[j,]=as.matrix(pred[[j]]$soj[[1]][k])
}
sojjson[[p]]=tmplistsoj
assign(paste0("forMSM",names(pred[[1]]$soj[[1]])[k]), sojjson[[p]])
p=p+1
}
}
sojvector=vector()
sojvector=ls(pattern = "forMSMSoj_.$")
sojvector=sub("forMSM","",ls(pattern = "^forMSMSoj_.$") )
sojvector=stringi::stri_sort(sojvector, numeric = TRUE)
names(sojjson)=sojvector
if (ci.json!="none" ) {
### Sojourn lci estimates #####
sojjson_lci=list()
if (length(covariates_list)<=1) {
p=1
for (k in na_states) {
tmplistsoj=matrix(nrow =1 , ncol= 1)
tmplistsoj[1,1]=as.matrix(pred[[1]]$soj_lci[[1]][1])
sojjson_lci[[1]]=tmplistsoj
assign(paste0("forMSM",names(pred[[1]]$soj_lci)[i]), tmplistsoj[[1]])
p=p+1
}
}
if (length(covariates_list)>1) {
p=1
for (k in na_states) {
tmplistsoj=matrix(nrow =length(covariates_list) , ncol= 1)
for (j in 1:length(covariates_list)) {
tmplistsoj[j,]=as.matrix(pred[[j]]$soj_lci[[1]][k])
}
sojjson_lci[[p]]=tmplistsoj
assign(paste0("forMSM",names(pred[[1]]$soj_lci[[1]])[k]), sojjson_lci[[p]])
p=p+1
}
}
sojvector_lci=vector()
sojvector_lci=ls(pattern = "^forMSMSoj.*lci$")
sojvector_lci=sub("forMSM","",ls(pattern = "^forMSMSoj.*lci$") )
sojvector_lci=stringi::stri_sort(sojvector_lci, numeric = TRUE)
names(sojjson_lci)=sojvector_lci
### Sojourn uci estimates #####
sojjson_uci=list()
if (length(covariates_list)<=1) {
p=1
for (k in na_states) {
tmplistsoj=matrix(nrow =1 , ncol= 1)
tmplistsoj[1,1]=as.matrix(pred[[1]]$soj_uci[[1]][1])
sojjson_uci[[1]]=tmplistsoj
assign(paste0("forMSM",names(pred[[1]]$soj_uci)[i]), tmplistsoj[[1]])
p=p+1
}
}
if (length(covariates_list)>1) {
p=1
for (k in na_states) {
tmplistsoj=matrix(nrow =length(covariates_list) , ncol= 1)
for (j in 1:length(covariates_list)) {
tmplistsoj[j,]=as.matrix(pred[[j]]$soj_uci[[1]][k])
}
sojjson_uci[[p]]=tmplistsoj
assign(paste0("forMSM",names(pred[[1]]$soj_uci[[1]])[k]), sojjson_uci[[p]])
p=p+1
}
}
sojvector_uci=vector()
sojvector_uci=ls(pattern = "^forMSMSoj.*uci$")
sojvector_uci=sub("forMSM","",ls(pattern = "^forMSMSoj.*uci$") )
sojvector_uci=stringi::stri_sort(sojvector_uci, numeric = TRUE)
names(sojjson_uci)=sojvector_uci
}
}
##########################################################################
### Next ################################################################
##############################################################################
if (pnext==TRUE) {
### Next Main estimates #####
nextjson=list()
if (length(covariates_list)<=1) {
for (k in 1:nstates) {
tmplistnext=matrix(nrow =1 , ncol= 1)
tmplistnext[1,1]=as.matrix(pred[[1]]$nextv[[1]][1])
nextjson[[1]]=tmplistnext
assign(paste0("forMSM",names(pred[[1]]$nextv)[i]), tmplistnext[[1]])
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
tmplistnext=matrix(nrow =length(covariates_list) , ncol= 1)
for (j in 1:length(covariates_list)) {
tmplistnext[j,]=as.matrix(pred[[j]]$nextv[[1]][k])
}
nextjson[[p]]=tmplistnext
assign(paste0("forMSM",names(pred[[1]]$nextv[[1]])[k]), nextjson[[p]])
p=p+1
}
}
nextvector=vector()
nextvector=ls(pattern = "forMSMNext_._to_.$")
nextvector=sub("forMSM","",ls(pattern = "^forMSMNext_._to_.$") )
nextvector=stringi::stri_sort(nextvector, numeric = TRUE)
names(nextjson)=nextvector
if (ci.json!="none" ) {
####### Next lci estimates ####################################
nextjson_lci=list()
if (length(covariates_list)<=1) {
for (k in 1:nstates) {
tmplistnext=matrix(nrow =1 , ncol= 1)
tmplistnext[1,1]=as.matrix(pred[[1]]$nextv_lci[[1]][1])
nextjson_lci[[1]]=tmplistnext
assign(paste0("forMSM",names(pred[[1]]$nextv_lci)[i]), tmplistnext[[1]])
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
tmplistnext=matrix(nrow =length(covariates_list) , ncol= 1)
for (j in 1:length(covariates_list)) {
tmplistnext[j,]=as.matrix(pred[[j]]$nextv_lci[[1]][k])
}
nextjson_lci[[p]]=tmplistnext
assign(paste0("forMSM",names(pred[[1]]$nextv_lci[[1]])[k]), nextjson_lci[[p]])
p=p+1
}
}
nextvector_lci=vector()
nextvector_lci=ls(pattern = "^forMSMNext.*lci$")
nextvector_lci=sub("forMSM","",ls(pattern = "^forMSMNext.*lci$") )
nextvector_lci=stringi::stri_sort(nextvector_lci, numeric = TRUE)
names(nextjson_lci)=nextvector_lci
####### Next uci estimates ####################################
nextjson_uci=list()
if (length(covariates_list)<=1) {
for (k in 1:nstates) {
tmplistnext=matrix(nrow =1 , ncol= 1)
tmplistnext[1,1]=as.matrix(pred[[1]]$nextv_uci[[1]][1])
nextjson_uci[[1]]=tmplistnext
assign(paste0("forMSM",names(pred[[1]]$nextv_uci)[i]), tmplistnext[[1]])
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
tmplistnext=matrix(nrow =length(covariates_list) , ncol= 1)
for (j in 1:length(covariates_list)) {
tmplistnext[j,]=as.matrix(pred[[j]]$nextv_uci[[1]][k])
}
nextjson_uci[[p]]=tmplistnext
assign(paste0("forMSM",names(pred[[1]]$nextv_uci[[1]])[k]), nextjson_uci[[p]])
p=p+1
}
}
nextvector_uci=vector()
nextvector_uci=ls(pattern = "^forMSMNext.*uci$")
nextvector_uci=sub("forMSM","",ls(pattern = "^forMSMNext.*uci$") )
nextvector_uci=stringi::stri_sort(nextvector_uci, numeric = TRUE)
names(nextjson_uci)=nextvector_uci
}
}
###################################################################################################
##########################################################################
### First ################################################################
##############################################################################
if (efpt==TRUE) {
### First Main estimates #####
firstjson=list()
if (length(covariates_list)<=1) {
for (k in 1:3) {
tmplistfirst=matrix(nrow =1 , ncol= 1)
tmplistfirst[1,1]=as.matrix(pred[[1]]$first[[1]][1])
firstjson[[1]]=tmplistfirst
assign(paste0("forMSM",names(pred[[1]]$first)[i]), tmplistfirst[[1]])
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:3) {
tmplistfirst=matrix(nrow =length(covariates_list) , ncol= 1)
for (j in 1:length(covariates_list)) {
tmplistfirst[j,]=as.matrix(pred[[j]]$first[[1]][k])
}
firstjson[[p]]=tmplistfirst
assign(paste0("forMSM",names(pred[[1]]$first[[1]])[k]), firstjson[[p]])
p=p+1
}
}
firstvector=vector()
firstvector=ls(pattern = "forMSMFirst_.$")
firstvector=sub("forMSM","",ls(pattern = "^forMSMFirst") )
firstvector=stringi::stri_sort(firstvector, numeric = TRUE)
names(firstjson)=firstvector
if (ci.json!="none" ) {
####### First lci estimates ####################################
firstjson_lci=list()
if (length(covariates_list)<=1) {
for (k in 1:nstates) {
tmplistfirst=matrix(nrow =1, ncol= 1)
tmplistfirst[1,1]=as.matrix(pred[[1]]$first_lci[[1]][1])
firstjson_lci[[1]]=tmplistfirst
assign(paste0("forMSM",names(pred[[1]]$first_lci)[i]), tmplistfirst[[1]])
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
tmplistfirst=matrix(nrow =length(covariates_list) , ncol= 1)
for (j in 1:length(covariates_list)) {
tmplistfirst[j,]=as.matrix(pred[[j]]$first_lci[[1]][k])
}
firstjson_lci[[p]]=tmplistfirst
assign(paste0("forMSM",names(pred[[1]]$first_lci[[1]])[k]), firstjson_lci[[p]])
p=p+1
}
}
firstvector_lci=vector()
firstvector_lci=ls(pattern = "^forMSMFirst.*lci$")
firstvector_lci=sub("forMSM","",ls(pattern ="^forMSMFirst.*lci$" ) )
firstvector_lci=stringi::stri_sort(firstvector_lci, numeric = TRUE)
names(firstjson_lci)=firstvector_lci
####### First uci estimates ####################################
firstjson_uci=list()
if (length(covariates_list)<=1) {
for (k in 1:nstates) {
tmplistfirst=matrix(nrow =1 , ncol= 1)
tmplistfirst[1,1]=as.matrix(pred[[1]]$first_uci[[1]][1])
firstjson_uci[[1]]=tmplistfirst
assign(paste0("forMSM",names(pred[[1]]$first_uci)[i]), tmplistfirst[[1]])
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
tmplistfirst=matrix(nrow =length(covariates_list) , ncol= 1)
for (j in 1:length(covariates_list)) {
tmplistfirst[j,]=as.matrix(pred[[j]]$first_uci[[1]][k])
}
firstjson_uci[[p]]=tmplistfirst
assign(paste0("forMSM",names(pred[[1]]$first_uci[[1]])[k]), firstjson_uci[[p]])
p=p+1
}
}
firstvector_uci=vector()
firstvector_uci=ls(pattern = "^forMSMFirst.*uci$")
firstvector_uci=sub("forMSM","",ls(pattern ="^forMSMFirst.*uci$" ) )
firstvector_uci=stringi::stri_sort(firstvector_uci, numeric = TRUE)
names(firstjson_uci)=firstvector_uci
}
}
################################################################################################
##########################################################################
### Number ################################################################
##############################################################################
if (envisits==TRUE) {
### Number Main estimates #####
numberjson=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
numberjson[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$number[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$number[[k]])[i]), numberjson[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplistnumber=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplistnumber[1,]=as.vector(pred[[1]]$number[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplistnumber[j,]=as.vector(pred[[j]]$number[[k]][,i])
}
numberjson[[p]]=tmplistnumber
assign(paste0("forMSM",names(pred[[1]]$number[[k]])[i]), numberjson[[p]])
p=p+1
}
}
}
numbervector=vector()
numbervector=ls(pattern = "forMSMNumber_._to_.$")
numbervector=sub("forMSM","",ls(pattern = "^forMSMNumber_._to_.$") )
numbervector=stringi::stri_sort(numbervector, numeric = TRUE)
names(numberjson)=numbervector
if (ci.json!="none" ) {
### Number lci estimates #####
numberjson_lci=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
numberjson_lci[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$number_lci[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$number_lci[[k]])[i]), numberjson_lci[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplistnumber=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplistnumber[1,]=as.vector(pred[[1]]$number_lci[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplistnumber[j,]=as.vector(pred[[j]]$number_lci[[k]][,i])
}
numberjson_lci[[p]]=tmplistnumber
assign(paste0("forMSM",names(pred[[1]]$number_lci[[k]])[i]), numberjson_lci[[p]])
p=p+1
}
}
}
numbervector_lci=vector()
numbervector_lci=ls(pattern = "^forMSMNumber.*lci$")
numbervector_lci=sub("forMSM","",ls(pattern = "^forMSMNumber.*lci$") )
numbervector_lci=stringi::stri_sort(numbervector_lci, numeric = TRUE)
names(numberjson_lci)=numbervector_lci
### Number uci estimates #####
numberjson_uci=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
numberjson_uci[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$number_uci[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$number_uci[[k]])[i]), numberjson_uci[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplistnumber=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplistnumber[1,]=as.vector(pred[[1]]$number_uci[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplistnumber[j,]=as.vector(pred[[j]]$number_uci[[k]][,i])
}
numberjson_uci[[p]]=tmplistnumber
assign(paste0("forMSM",names(pred[[1]]$number_uci[[k]])[i]), numberjson_uci[[p]])
p=p+1
}
}
}
numbervector_uci=vector()
numbervector_uci=ls(pattern = "^forMSMNumber.*uci$")
numbervector_uci=sub("forMSM","",ls(pattern = "^forMSMNumber.*uci$") )
numbervector_uci=stringi::stri_sort(numbervector_uci, numeric = TRUE)
names(numberjson_uci)=numbervector_uci
}
}
######################################################################################################
##########################################################################
### Visit ################################################################
##############################################################################
if (visit==TRUE) {
### Visit Main estimates #####
vjson=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
vjson[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$visit[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$visit[[k]])[i]), vjson[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplistvisit=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplistvisit[1,]=as.vector(pred[[1]]$visit[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplistvisit[j,]=as.vector(pred[[j]]$visit[[k]][,i])
}
vjson[[p]]=tmplistvisit
assign(paste0("forMSM",names(pred[[1]]$visit[[k]])[i]), vjson[[p]])
p=p+1
}
}
}
visvector=vector()
visvector=ls(pattern = "forMSMVisit_._to_.$")
visvector=sub("forMSM","",ls(pattern = "^forMSMVisit_._to_.$") )
visvector=stringi::stri_sort(visvector, numeric = TRUE)
names(vjson)=visvector
if (ci.json!="none" ) {
### Visit lci estimates #####
vjson_lci=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
vjson_lci[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$visit_lci[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$visit_lci[[k]])[i]), vjson_lci[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplistvisit=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplistvisit[1,]=as.vector(pred[[1]]$visit_lci[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplistvisit[j,]=as.vector(pred[[j]]$visit_lci[[k]][,i])
}
vjson_lci[[p]]=tmplistvisit
assign(paste0("forMSM",names(pred[[1]]$visit_lci[[k]])[i]), vjson_lci[[p]])
p=p+1
}
}
}
visvector_lci=vector()
visvector_lci=ls(pattern = "^forMSMVisit.*lci$")
visvector_lci=sub("forMSM","",ls(pattern ="^forMSMVisit.*lci$") )
visvector_lci=stringi::stri_sort(visvector_lci, numeric = TRUE)
names(vjson_lci)=visvector_lci[1:(p-1)]
### Visit uci estimates #####
vjson_uci=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
vjson_uci[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$visit_uci[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$visit_uci[[k]])[i]), vjson_uci[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplistvisit=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplistvisit[1,]=as.vector(pred[[1]]$visit_uci[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplistvisit[j,]=as.vector(pred[[j]]$visit_uci[[k]][,i])
}
vjson_uci[[p]]=tmplistvisit
assign(paste0("forMSM",names(pred[[1]]$visit_uci[[k]])[i]), vjson_uci[[p]])
p=p+1
}
}
}
visvector_uci=vector()
visvector_uci=ls(pattern = "^forMSMVisit.*uci$")
visvector_uci=sub("forMSM","",ls(pattern ="^forMSMVisit.*uci$") )
visvector_uci=stringi::stri_sort(visvector_uci, numeric = TRUE)
names(vjson_uci)=visvector_uci[1:(p-1)]
}
}
######################################################################################################
##########################################################################################
### vartime###
timejson=list()
#timematrix=matrix(nrow=1, ncol=length(vartime), NA)
#timematrix[1,]=vartime
timejson[[1]]=vartime
######################################################################################################
### At list ###
atlistjson=list()
if (length(covariates_list)>=1) {
atlistmatrix=matrix(nrow=1, ncol=length(covariates_list), NA)
for (j in 1:length(covariates_list) ) {
name_paste=list()
for (i in 1:length(covariates_list[[j]]) ) {
name_paste[[i]]= paste0(names(covariates_list[[j]])[i]," ", as.character(covariates_list[[j]][i]) )
}
atlistmatrix[1,j]= paste(unlist(name_paste,recursive=FALSE), collapse = ' ')
}
}
if (length(covariates_list)==0) {
atlistmatrix=matrix(nrow=1, ncol=1, NA)
name_paste= "all"
atlistmatrix[1,1]= name_paste
}
atlistjson[[1]]=as.vector(atlistmatrix)
######################################################################################################
final_list=list(timevar=timejson, Nats=length(covariates_list) ,
atlist=atlistjson, tmat=tmatlist, intermediate_states=intermediate_list, Ntransitions= ntransitions,
pjson,pjson_lci,pjson_uci,
hjson,hjson_lci,hjson_uci,
losjson,losjson_lci,losjson_uci,
sojjson,sojjson_lci,sojjson_uci,
nextjson,nextjson_lci,nextjson_uci,
firstjson, firstjson_lci, firstjson_uci,
numberjson,numberjson_lci,numberjson_uci,
vjson,vjson_lci,vjson_uci
)
final_list$timevar=as.vector(final_list$timevar)
# final_list$atlistjson=as.vector(final_list$atlistjson[[1]])
final_unlist= unlist(final_list, recursive=FALSE)
exportJson <- RJSONIO::toJSON(final_unlist, pretty = TRUE,force = TRUE, flatten=TRUE, na='string')
#exportJson
wd=getwd()
if (is.null(jsonpath) ) {
write(exportJson , paste0(wd ,"/", name ) )
}
if (!is.null(jsonpath)) {
if (nchar(jsonpath)==0) {
write(exportJson , paste0(wd ,"/", name ) )
}
if (nchar(jsonpath)!=0) {
write(exportJson, paste0(jsonpath ,"/", name ) )
}
}
rm(list=ls(pattern="^forMSM"))
final_list
}
nskourlis/MSMplus documentation built on Dec. 7, 2023, 4:53 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions msmjson
Documented in msmjson
#' @title msmjson
#' @description Function msmjson can receive models from msm package.
#' It then uses functions from msm package internally to estimate multi-state model
#' measures such as transition probabilities, transition intensity rates, length of stay, probability of visiting each state
#' probability each state is next, sojoutn times and confidence intervals of the
#' estimations. Function flexjson then take these results and reshapes them so that they can
#' be fed to MSMplus properly as a json file.
#' @param msm.model A fitted multi-state model, as returned by msm.
#' @param vartime The time points to estimate the transition probabilities for, by default one unit.
#' @param mat.init A transition intensity matrix. Default: NULL
#' @param totlos Estimate the total length of stay in each state for the time points
#' specified by vartime argument, Default: "FALSE"
#' @param visit Estimate the probability of visiting each state for the time points
#' specified by vartime argument, Default: "FALSE"
#' @param sojourn Estimate the mean sojourn time in each state, Default: "FALSE"
#' @param pnext Estimate the probability that each state is next, Default: "FALSE"
#' @param efpt Estimate the expected first passage time by each state, Default: "FALSE"
#' @param envisits Estimate the number of visits in each state , Default: "FALSE"
#' @param ci.json Estimate confidence intervals, Default: 'normal'
#' @param cl.json Specify confidence level, Default: 0.95
#' @param B.json Number of simulations from the normal asymptotic distribution used to calculate variances.
#' Decrease for greater speed at the expense of accuracy, Default: 50
#' @param cores.json Number of cores to use for bootstrapping using parallel processing. See boot.msm for more details., Default: NULL
#' @param piecewise.times.json Times at which piecewise-constant intensities change.
#' See pmatrix.piecewise.msm for how to specify this. This is only required for time-inhomogeneous models
#' specified using explicit time-dependent covariates, and should not be used for models specified using "pci"., Default: NULL
#' @param piecewise.covariates.json Covariates on which the piecewise-constant intensities depend. See pmatrix.piecewise.msm for how to specify this., Default: NULL
#' @param num.integ.json Use numerical integration instead of analytic solution (see below).
#' @param covariates_list The user can specify different covariate patterns
#' for which predictions will be made, Default: list()
#' @param jsonpath specify the path of the folder that the json file should be saved, Default: "" saves the json file to the current working directory
#' @param name Specify the name of the output json file, Default: 'predictions.json'
#' @return returns a list of objects: the time variable
#' the number of covariate patterns, the names of covariate patterns, the transition matrix,
#' the number of transitions, the transition probabilities, transition intensity rates, length of stay, probability of visiting each state
#' probability each state is next, sojoutn times and confidence intervals of the
#' estimations(if estimated).
#' @details DETAILS
#' @examples
#' \dontrun{
#'
#' #Example
#'
#' # Multi-state model analysis: Using msmjson function together with msm package
#'
#' library("MSMplus")
#' library("msm")
#' library("dplyr")
#'
#' options(scipen = 999,"digits"=10)
#'
#' head(cav)
#'
#' ### Renaming variable PTNUM to id
#'
#' cav$id=cav$PTNUM
#'
#' ### Defining the transition matrix
#'
#' tmat=matrix(NA,nrow=4,ncol=4)
#' tmat[1,2]=1; tmat[1,4]=2; tmat[2,1]=3; tmat[2,3]=4
#' tmat[2,4]=5; tmat[3,2]=6; tmat[3,4]=7
#'
#' ### We can now call function msboxes_R
#'
#' results3_days=MSMplus::msboxes_R(data=cav,id= cav$id, yb=c(0.3,0.5,0.6,0.75), msm=TRUE,
#' xb=c(0.5,0.2,0.7,0.3),boxwidth=0.1,boxheight=0.1,
#' tmatrix= tmat, vartime=seq(0,10,by=1),scale=1,
#' jsonpath="", name="msboxes_cav_R.json" )
#'
#'
#' ### Defining the transition matrix with initial values under an initial assumption
#'
#' #0 for transitions not allowed, initial values for rest of transitions under a rationale ##
#'
#' Q<- rbind(c(0,0.25,0,0.25),c(0.166,0,0.166,0.166),c(0,0.25,0,0.25),c(0,0,0,0))
#'
#'
#' ### Getting initial Q matrix in a default way- Feed the hand made matrix
#'
#' q.crude<- crudeinits.msm(state~years, id,data=cav, qmatrix=Q)
#'
#'
#' ### Apply the msm model
#'
#' cavsex.msm<- msm(state~years, covariates=~sex, id,data=cav,qmatrix=q.crude, deathexact = 4, control=list(trace=1,REPORT=1))
#' summary(cavsex.msm)
#'
#'
#' ### Prediction for different covariate patterns (males and females)
#'
#' results <- MSMplus::msmjson(msm.model=cavsex.msm, vartime=seq(1,3,1), mat.init=q.crude,
#' totlos=TRUE, visit=TRUE, sojourn=TRUE, pnext=TRUE, efpt=TRUE, envisits=TRUE,
#' ci.json="normal", cl.json=0.95, B.json=100,
#' cores.json=NULL,piecewise.times.json=NULL, piecewise.covariates.json=NULL,num.integ.json=FALSE,
#' covariates_list=list(list(sex = 1),list(sex = 0)),
#' jsonpath="",
#' name="predictions_cav_R.json" )
#'
#' }
#' @seealso
#' \code{\link[stringi]{stri_sort}}
#' @rdname msmjson
#' @export
#' @importFrom stringi stri_sort
msmjson <- function(msm.model, vartime=seq(1,1,1), mat.init,
totlos=FALSE, visit=FALSE, sojourn=FALSE, pnext=FALSE, efpt=FALSE, envisits=FALSE,
ci.json="normal", cl.json=0.95, B.json=50,
cores.json=NULL,piecewise.times.json=NULL, piecewise.covariates.json=NULL,num.integ.json=FALSE,
covariates_list=list(), jsonpath="",name="predictions_R.json" ) {
### Default for jsonpath is user's home directory ####
options(scipen = 999,"digits"=10)
#
# if (is.null(covariate_patterns)) {covariates_list=list()}
#
# else if ( !is.null(covariate_patterns) ) {
# covariates_list=list()
# for (i in 1:nrow(covariate_patterns)) {covariates_list[[i]]=covariate_patterns[i,]}
# }
################################################################
nstates=ncol(mat.init)
p=1
trmat=matrix(nrow=nstates, ncol=nstates, NA)
mat=mat.init
mat[is.na(mat)] <- 0
ntransitions=length(mat[which(mat>0)])
nstates=ncol(mat)
states=c(seq(1:nstates))
trmat=matrix(nrow=nstates, ncol=nstates, NA)
for (i in 1:nstates) {
for (k in 1:nstates) {
if (mat[i,k]>0) {
trmat[i,k]=as.integer(p)
p=p+1
}
else if (mat[i,k]<=0) {
trmat[i,k]=NA
}
}
}
tmatlist=list()
tmatlist[[1]]=trmat
tr_start_state=vector()
for (k in 1:ntransitions) {
tr_start_state[k]=which(trmat == k, arr.ind = TRUE)[1,1]
}
tr_end_state=vector()
for (k in 1:ntransitions) {
tr_end_state[k]=which(trmat == k, arr.ind = TRUE)[1,2]
}
state_kind=vector()
for (i in 1:nstates) {
if (length(which(mat[,i]==0))==nrow(mat)) {state_kind[i]="Starting"}
else if (length(which(mat[i,]==0))==ncol(mat)) {state_kind[i]="Absorbing"}
else {state_kind[i]="Intermediate"}
}
st_states=which(state_kind=="Starting")
na_states=which(state_kind=="Intermediate")
ab_states=which(state_kind=="Absorbing")
intermediate_list=list()
intermediate_list[[1]]=unique(tr_start_state)
Ntransitions=length(mat[which(mat>0)])
###################################################################
# library("msm")
# library("stringi")
# library("RJSONIO")
pred=list()
#if (length(vartime)==1) {vartime=t(vartime)}
### First case, overall predictions
if (length(covariates_list) == 0) {
################################################################################################################
### Create all the lists to be output even if they will remain vacant ###
loslist=list()
loslist_lci=list()
loslist_uci=list()
sojlist=list()
sojlist_lci=list()
sojlist_uci=list()
nextlist=list()
nextlist_lci=list()
nextlist_uci=list()
firstlist=list()
firstlist_lci=list()
firstlist_uci=list()
numberlist=list()
numberlist_lci=list()
numberlist_uci=list()
vlist=list()
vlist_lci=list()
vlist_uci=list()
trans_lci=vector()
trans_uci=vector()
#################################################################################################################
###Probability ###
mat=mat.init
mat[is.na(mat)] <- 0
ntransitions=length(mat[which(mat>0)])
nstates=ncol(mat)
states=c(seq(1:nstates))
##################### Main estimates#####################################################################
p_array= array(dim=c(length(vartime),nstates+1,nstates),NA)
p_array_lci= array(dim=c(length(vartime),nstates+1,nstates),NA)
p_array_uci= array(dim=c(length(vartime),nstates+1,nstates),NA)
for (i in 1:length(vartime)) {
p.msm <- pmatrix.msm(msm.model,t=vartime[i], t1=0, ci=ci.json , cl=cl.json, B=B.json,
cores=cores.json )
if (ci.json=="none" ) {
p_array[i,,]=t(cbind(p.msm,vartime=vartime[i] ))
}
if (ci.json!="none" ) {
p_array[i,,]=t(cbind(p.msm$estimates,vartime=vartime[i] ))
p_array_lci[i,,]=t(cbind(p.msm$L,vartime=vartime[i] ))
p_array_uci[i,,]=t(cbind(p.msm$U,vartime=vartime[i] ))
}
}
plist=list()
for (k in states) {
plist[[k]]=p_array[,,k]
plist[[k]]=as.data.frame(plist[[k]])
}
if (length(vartime)==1) {
for (k in states) {
plist[[k]]=as.data.frame(t(plist[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(plist[[k]])[j]=paste0("P_",k,"_to_",j)
}
colnames(plist[[k]])[nstates+1] ="vartime"
}
plist
plist_lci=list()
plist_uci=list()
if (ci.json!="none" ) {
##################### Probabilities lci #####################################################################
for (k in states) {
plist_lci[[k]]=p_array_lci[,,k]
plist_lci[[k]]=as.data.frame(plist_lci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
plist_lci[[k]]=as.data.frame(t(plist_lci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(plist_lci[[k]])[j]=paste0("P_",k,"_to_",j,"_lci")
}
colnames(plist_lci[[k]])[nstates+1] ="vartime"
}
plist_lci
##################### Probabilities uci #####################################################################
plist_uci=list()
for (k in states) {
plist_uci[[k]]=p_array_uci[,,k]
plist_uci[[k]]=as.data.frame(plist_uci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
plist_uci[[k]]=as.data.frame(t(plist_uci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(plist_uci[[k]])[j]=paste0("P_",k,"_to_",j,"_uci")
}
colnames(plist_uci[[k]])[nstates+1] ="vartime"
}
plist_uci
}
#################################################################################################################
###Transitions values for the period we want (of course they are constant)###
##################### Main estimates#####################################################################
qmsm=qmatrix.msm(msm.model,ci=ci.json , cl=cl.json,
B=B.json, cores=cores.json)
if (ci.json!="none" ) {
trans=matrix(nrow=length(vartime), ncol=length(qmsm$estimates[which(qmsm$estimates>0)]), NA)
for (i in 1:length(vartime)){
trans[i,]= qmsm$estimates[which(qmsm$estimates>0)]
}
trans=as.data.frame(trans)
for (i in 1:length(qmsm$estimates[which(qmsm$estimates>0)]) ){
colnames(trans)[i]<-paste0("Haz_",tr_start_state[i],"_to_",tr_end_state[i])
}
}
if (ci.json=="none" ) {
trans=matrix(nrow=length(vartime), ncol=length(qmsm[which(qmsm>0)]), NA)
for (i in 1:length(vartime)){
trans[i,]= qmsm[which(qmsm>0)]
}
trans=as.data.frame(trans)
for (i in 1:length(qmsm[which(qmsm>0)]) ){
colnames(trans)[i]<-paste0("Haz_",tr_start_state[i],"_to_",tr_end_state[i])
}
}
trans$vartime=vartime
trans
##################### transitions lci and uci #####################################################################
if (ci.json!="none" ) {
trans_lci=matrix(nrow=length(vartime), ncol=length(qmsm$estimates[which(qmsm$L>0)]), NA)
for (i in 1:length(vartime)){
trans_lci[i,]= qmsm$L[which(qmsm$L>0)]
}
trans_lci=as.data.frame(trans_lci)
for (i in 1:length(qmsm$L[which(qmsm$L>0)]) ){
colnames(trans_lci)[i]<- paste0("Haz_",tr_start_state[i],"_to_",tr_end_state[i],"_lci")
}
trans_lci$vartime=vartime
trans_lci
trans_uci=matrix(nrow=length(vartime), ncol=length(qmsm$estimates[which(qmsm$L>0)]), NA)
for (i in 1:length(vartime)){
trans_uci[i,]= qmsm$L[which(qmsm$L>0)]
}
trans_uci=as.data.frame(trans_uci)
for (i in 1:length(qmsm$L[which(qmsm$L>0)]) ){
colnames(trans_uci)[i]<-paste0("Haz_",tr_start_state[i],"_to_",tr_end_state[i],"_uci")
}
trans_uci$vartime=vartime
trans_uci
}
#################################################################################################################
###Los ###
if (totlos==TRUE) {
los_array= array(dim=c(length(vartime),nstates+1,nstates),NA)
los_array_lci= array(dim=c(length(vartime),nstates+1,nstates),NA)
los_array_uci= array(dim=c(length(vartime),nstates+1,nstates),NA)
for (j in states) {
for (i in 1:length(vartime)) {
los.msm <- totlos.msm(msm.model, start=j, fromt=0, tot= vartime[i],ci=ci.json,cl=cl.json,
piecewise.times=piecewise.times.json, piecewise.covariates=piecewise.covariates.json,
num.integ=num.integ.json, B=B.json,cores=cores.json)
if (ci.json=="none" ) { los_array[i,,j]=cbind(t(los.msm),vartime=vartime[i] ) }
if (ci.json!="none" ) {
los_array[i,,j]=cbind(t(los.msm[1,]),vartime=vartime[i] )
los_array_lci[i,,j]=cbind(t(round(los.msm[2,],5)),vartime=vartime[i] )
los_array_uci[i,,j]=cbind(t(round(los.msm[3,],5)),vartime=vartime[i] )
}
}
}
loslist=list()
for (k in states) {
loslist[[k]]=los_array[,,k]
loslist[[k]]=as.data.frame(loslist[[k]])
}
if (length(vartime)==1) {
for (k in states) {
loslist[[k]]=as.data.frame(t(loslist[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(loslist[[k]])[j]=paste0("Los_",k,"_to_",j)
}
colnames(loslist[[k]])[nstates+1] ="vartime"
}
loslist
##################### los lci and uci #####################################################################
loslist_lci=list()
loslist_uci=list()
if (ci.json!="none" ) {
###lci
for (k in states) {
loslist_lci[[k]]=los_array_lci[,,k]
loslist_lci[[k]]=as.data.frame(loslist_lci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
loslist_lci[[k]]=as.data.frame(t(loslist_lci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(loslist_lci[[k]])[j]=paste0("Los_",k,"_to_",j,"_lci")
}
colnames(loslist_lci[[k]])[nstates+1] ="vartime"
}
loslist_lci
###uci
for (k in states) {
loslist_uci[[k]]=los_array_uci[,,k]
loslist_uci[[k]]=as.data.frame(loslist_uci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
loslist_uci[[k]]=as.data.frame(t(loslist_uci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(loslist_uci[[k]])[j]=paste0("Los_",k,"_to_",j,"_uci")
}
colnames(loslist_uci[[k]])[nstates+1] ="vartime"
}
loslist_uci
}
}
#################################################################################################################
###Sojourn ###
if (sojourn==TRUE) {
sojlist_=list()
sojlist_lci=list()
sojlist_uci=list()
soj.msm=vector()
if (ci.json!="none") {
soj.msm=sojourn.msm(msm.model,ci=ci.json,cl=cl.json, B=B.json)
sojnames= rownames(soj.msm)
sojlist=list()
sojlist[[1]]=cbind.data.frame(t( soj.msm$estimates))
names(sojlist[[1]])[1:(ncol(sojlist[[1]]))]<-paste0("Soj_",sub("State ","",sojnames))
sojlist
sojlist_lci[[1]]=cbind.data.frame(t(soj.msm$L))
sojlist_uci[[1]]=cbind.data.frame(t(soj.msm$U))
names(sojlist_lci[[1]])[1:(ncol(sojlist_lci[[1]]))]<-paste0("Soj_",sub("State ","",sojnames),"_lci")
names(sojlist_uci[[1]])[1:(ncol(sojlist_uci[[1]]))]<-paste0("Soj_",sub("State ","",sojnames),"_uci")
}
else if (ci.json =="none") {
soj.msm=sojourn.msm(msm.model,ci=ci.json,cl=cl.json, B=B.json)
sojlist=list()
sojlist[[1]]=cbind.data.frame(t( soj.msm$estimates))
sojnames=c(1:ncol(sojlist[[1]]))
names(sojlist[[1]])[1:(ncol(sojlist[[1]]))]<-paste0("Soj_",sub("State ","",sojnames))
sojlist
}
}
#################################################################################################################
###Next ###
########################### Main estimates ####################################################
if (pnext==TRUE) {
next_array= array(dim=c(1,nstates,nstates),NA)
next.msm <- pnext.msm(msm.model,ci=ci.json,cl=cl.json, B=B.json,cores=cores.json)
next_array[1,,]=t(cbind.data.frame(next.msm$estimates ))
nextlist=list()
for (k in states) {
nextlist[[k]]=next_array[,,k]
nextlist[[k]]=as.data.frame(t(nextlist[[k]]))
}
# if (length(vartime)==1) {
# for (k in states) {
# nextlist[[k]]=as.data.frame(t(nextlist[[k]]))
# }
# }
for (k in states) {
for (j in states) {
colnames(nextlist[[k]])[j]=paste0("Next_",k,"_to_",j)
}
}
nextlist
########################### next uci , lci ####################################################
nextlist_lci=list()
nextlist_uci=list()
if (ci.json!="none" ) {
###lci
next_array_lci= array(dim=c(1,nstates,nstates),NA)
next_array_lci[1,,]=t(cbind.data.frame(next.msm$L ))
for (k in states) {
nextlist_lci[[k]]=next_array_lci[,,k]
nextlist_lci[[k]]=as.data.frame(t(nextlist_lci[[k]]))
}
# if (length(vartime)==1) {
# for (k in states) {
# nextlist_lci[[k]]=as.data.frame(t(nextlist_lci[[k]]))
# }
# }
for (k in states) {
for (j in states) {
colnames(nextlist_lci[[k]])[j]=paste0("Next_",k,"_to_",j,"_lci")
}
}
nextlist_lci
### uci
next_array_uci= array(dim=c(1,nstates,nstates),NA)
next_array_uci[1,,]=t(cbind.data.frame(next.msm$U ))
for (k in states) {
nextlist_uci[[k]]=next_array_uci[,,k]
nextlist_uci[[k]]=as.data.frame(t(nextlist_uci[[k]]))
}
# if (length(vartime)==1) {
# for (k in states) {
# nextlist_uci[[k]]=as.data.frame(t(nextlist_uci[[k]]))
# }
# }
#
for (k in states) {
for (j in states) {
colnames(nextlist_uci[[k]])[j]=paste0("Next_",k,"_to_",j,"_uci")
}
}
nextlist_uci
}
}
###############################################################################
##First
if (efpt==TRUE) {
first.msm=vector()
first.msm= efpt.msm(msm.model, start="all", tostate=states,
ci=ci.json,cl=cl.json, B=B.json,cores=cores.json)
firstlist=list()
if (ci.json=="none" ) {
firstlist[[1]]=cbind.data.frame(t( first.msm))
}
if (ci.json!="none" ) {
firstlist[[1]]=cbind.data.frame(t( first.msm[1,]))
}
for (i in 1:nstates) {
names( firstlist[[1]])[i]<-paste0("First_",i)
}
firstlist_lci=list()
firstlist_uci=list()
if (ci.json!="none" ) {
firstlist_lci[[1]]=cbind.data.frame(t(first.msm[2,]))
firstlist_uci[[1]]=cbind.data.frame(t(first.msm[3,]))
for (i in 1:nstates) {
names( firstlist_lci[[1]])[i]<-paste0("First_",i,"_lci")
names( firstlist_uci[[1]])[i]<-paste0("First_",i,"_uci")
}
}
}
##########################################################################################
##Number
if (envisits==TRUE) {
############Main estimates ########################
number_array= array(dim=c(length(vartime),nstates+1,nstates),NA)
number_array_lci= array(dim=c(length(vartime),nstates+1,nstates),NA)
number_array_uci= array(dim=c(length(vartime),nstates+1,nstates),NA)
for (j in states) {
for (i in 1:length(vartime)) {
number.msm <- envisits.msm(msm.model, start=j, fromt=0, tot= vartime[i],
piecewise.times=piecewise.times.json, piecewise.covariates=piecewise.covariates.json,
num.integ=num.integ.json,
ci=ci.json, cl=cl.json, B=B.json,
cores=cores.json)
if (ci.json=="none" ) { number_array[i,,j]=cbind(t(number.msm),vartime=vartime[i] ) }
if (ci.json!="none" ) {
number_array[i,,j]=cbind(t(round(number.msm[1,],5)),vartime=vartime[i] )
number_array_lci[i,,j]=cbind(t(round(number.msm[2,],5)),vartime=vartime[i] )
number_array_uci[i,,j]=cbind(t(round(number.msm[3,],5)),vartime=vartime[i] )
}
}
}
numberlist=list()
for (k in states) {
numberlist[[k]]=number_array[,,k]
numberlist[[k]]=as.data.frame(numberlist[[k]])
}
if (length(vartime)==1) {
for (k in states) {
numberlist[[k]]=as.data.frame(t(numberlist[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(numberlist[[k]])[j]=paste0("Number_",k,"_to_",j)
}
colnames(numberlist[[k]])[nstates+1] ="vartime"
}
numberlist
numberlist_lci=list()
numberlist_uci=list()
if (ci.json!="none" ) {
############ lci estimates ########################
for (k in states) {
numberlist_lci[[k]]=number_array_lci[,,k]
numberlist_lci[[k]]=as.data.frame(numberlist_lci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
numberlist_lci[[k]]=as.data.frame(t(numberlist_lci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(numberlist_lci[[k]])[j]=paste0("Number_",k,"_to_",j,"_lci")
}
colnames(numberlist_lci[[k]])[nstates+1] ="vartime"
}
numberlist_lci
############ uci estimates ########################
for (k in states) {
numberlist_uci[[k]]=number_array_uci[,,k]
numberlist_uci[[k]]=as.data.frame(numberlist_uci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
numberlist_uci[[k]]=as.data.frame(t(numberlist_uci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(numberlist_uci[[k]])[j]=paste0("Number_",k,"_to_",j,"_uci")
}
colnames(numberlist_uci[[k]])[nstates+1] ="vartime"
}
numberlist_uci
}
}
###########################################################################################################
##Visit
if (visit==TRUE) {
visit.msm=vector()
######## What should we put in start? 1 or all states
visit_array= array(dim=c(length(vartime),nstates+1,nstates),NA)
visit_array_lci= array(dim=c(length(vartime),nstates+1,nstates),NA)
visit_array_uci= array(dim=c(length(vartime),nstates+1,nstates),NA)
for (i in 1:length(vartime)) {
visit.msm <- ppass.msm(x=msm.model, qmatrix=NULL, tot=vartime[i], start="all",
piecewise.times=piecewise.times.json, piecewise.covariates=piecewise.covariates.json,
ci=ci.json, cl=cl.json, B=B.json,
cores=cores.json)
if (ci.json=="none" ) {visit_array[i,,]=t(cbind(visit.msm,vartime=vartime[i] )) }
if (ci.json!="none" ) {
visit_array[i,,]=t(cbind(visit.msm$estimates,vartime=vartime[i] ))
visit_array_lci[i,,]=t(cbind(visit.msm$L,vartime=vartime[i] ))
visit_array_uci[i,,]=t(cbind(visit.msm$U,vartime=vartime[i] ))
}
}
####################### Main estimates ########################
vlist=list()
for (k in states) {
vlist[[k]]=visit_array[,,k]
vlist[[k]]=as.data.frame(vlist[[k]])
}
if (length(vartime)==1) {
for (k in states) {
vlist[[k]]=as.data.frame(t(vlist[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(vlist[[k]])[j]=paste0("Visit_",k,"_to_",j)
}
colnames(vlist[[k]])[nstates+1] ="vartime"
}
vlist
vlist_lci=list()
vlist_uci=list()
if (ci.json!="none" ) {
####################### Lci estimates ########################
for (k in states) {
vlist_lci[[k]]=visit_array_lci[,,k]
vlist_lci[[k]]=as.data.frame(vlist_lci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
vlist_lci[[k]]=as.data.frame(t(vlist_lci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(vlist_lci[[k]])[j]=paste0("Visit_",k,"_to_",j,"_lci")
}
colnames(vlist_lci[[k]])[nstates+1] ="vartime"
}
vlist_lci
####################### Uci estimates ########################
for (k in states) {
vlist_uci[[k]]=visit_array_uci[,,k]
vlist_uci[[k]]=as.data.frame(vlist_uci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
vlist_uci[[k]]=as.data.frame(t(vlist_uci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(vlist_uci[[k]])[j]=paste0("Visit_",k,"_to_",j,"_uci")
}
colnames(vlist_uci[[k]])[nstates+1] ="vartime"
}
vlist_uci
}
}
###########################################################################################
# ###Ratio of tr. intensities
# mat
#
# p=1
# qratio=matrix(nrow=1, ncol=length(which(mat>0))^2,NA)
# names.qratio=vector()
#
# for (i in 1:nstates) {
# for (j in 1:nstates) {
# for (k in 1:nstates) {
# for (f in 1:nstates) {
#
# if (mat[k,f]==0 | mat[i,j]==0 | k==f | i==j) {next}
# else if (mat[k,f]!=0 & mat[i,j]!=0 & k!=f & i!=j ) {
# qratio[1,p]=qratio.msm(msm.model, ind1=c(i,j), ind2=c(k,f), covariates = covariates_list[[g]])[1]
# names.qratio[p]=paste0("Qratio_t1_",i,".",j,"_t2_",k,".",f)
# p=p+1
# }
# }
# }
# }
# }
#
# qratio=as.data.frame(qratio)
# names(qratio)=names.qratio
#
# qratiolist=list()
# qratiolist[[1]]=cbind.data.frame(qratio,cov_factor=rep(paste0(names(covariates_list[[g]]),covariates_list[[g]]),1))
##############################################################################################
pred[[1]]= list(probs=plist,probs_lci=plist_lci,probs_uci=plist_uci,
trans=trans, trans_lci=trans_lci,trans_uci=trans_uci,
los=loslist,los_lci=loslist_lci,los_uci=loslist_uci,
soj=sojlist,soj_lci=sojlist_lci,soj_uci=sojlist_uci,
nextv=nextlist,nextv_lci=nextlist_lci,nextv_uci=nextlist_uci,
first=firstlist,first_lci=firstlist_lci,first_uci=firstlist_uci,
number=numberlist,number_lci=numberlist_lci,number_uci=numberlist_uci,
visit=vlist,visit_lci=vlist_lci, visit_uci=vlist_uci
)
}
if (length(covariates_list) != 0) {
for (g in 1:length(covariates_list)) {
#g=1
################################################################################################################
### Create all the lists to be output even if they will remain vacant ###
loslist=list()
loslist_lci=list()
loslist_uci=list()
sojlist=list()
sojlist_lci=list()
sojlist_uci=list()
nextlist=list()
nextlist_lci=list()
nextlist_uci=list()
firstlist=list()
firstlist_lci=list()
firstlist_uci=list()
numberlist=list()
numberlist_lci=list()
numberlist_uci=list()
vlist=list()
vlist_lci=list()
vlist_uci=list()
trans_lci=vector()
trans_uci=vector()
#################################################################################################################
###Probability ###
mat=mat.init
mat[is.na(mat)] <- 0
ntransitions=length(mat[which(mat>0)])
nstates=ncol(mat)
states=c(seq(1:nstates))
##################### Main estimates#####################################################################
p_array= array(dim=c(length(vartime),nstates+1,nstates),NA)
p_array_lci= array(dim=c(length(vartime),nstates+1,nstates),NA)
p_array_uci= array(dim=c(length(vartime),nstates+1,nstates),NA)
for (i in 1:length(vartime)) {
p.msm <- pmatrix.msm(msm.model,t=vartime[i], t1=0, ci=ci.json , cl=cl.json, B=B.json,
cores=cores.json ,covariates= covariates_list[[g]])
if (ci.json=="none" ) {
p_array[i,,]=t(cbind(p.msm,vartime=vartime[i] ))
}
if (ci.json!="none" ) {
p_array[i,,]=t(cbind(p.msm$estimates,vartime=vartime[i] ))
p_array_lci[i,,]=t(cbind(p.msm$L,vartime=vartime[i] ))
p_array_uci[i,,]=t(cbind(p.msm$U,vartime=vartime[i] ))
}
}
plist=list()
for (k in states) {
plist[[k]]=p_array[,,k]
plist[[k]]=as.data.frame(plist[[k]])
}
if (length(vartime)==1) {
for (k in states) {
plist[[k]]=as.data.frame(t(plist[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(plist[[k]])[j]=paste0("P_",k,"_to_",j)
}
colnames(plist[[k]])[nstates+1] ="vartime"
}
plist
plist_lci=list()
plist_uci=list()
if (ci.json!="none" ) {
##################### Probabilities lci #####################################################################
for (k in states) {
plist_lci[[k]]=p_array_lci[,,k]
plist_lci[[k]]=as.data.frame(plist_lci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
plist_lci[[k]]=as.data.frame(t(plist_lci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(plist_lci[[k]])[j]=paste0("P_",k,"_to_",j,"_lci")
}
colnames(plist_lci[[k]])[nstates+1] ="vartime"
}
plist_lci
##################### Probabilities uci #####################################################################
plist_uci=list()
for (k in states) {
plist_uci[[k]]=p_array_uci[,,k]
plist_uci[[k]]=as.data.frame(plist_uci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
plist_uci[[k]]=as.data.frame(t(plist_uci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(plist_uci[[k]])[j]=paste0("P_",k,"_to_",j,"_uci")
}
colnames(plist_uci[[k]])[nstates+1] ="vartime"
}
plist_uci
}
#################################################################################################################
###Transitions values for the period we want (of course they are constant)###
##################### Main estimates#####################################################################
qmsm=qmatrix.msm(msm.model,covariates=covariates_list[[g]],ci=ci.json , cl=cl.json,
B=B.json, cores=cores.json)
if (ci.json!="none" ) {
trans=matrix(nrow=length(vartime), ncol=length(qmsm$estimates[which(qmsm$estimates>0)]), NA)
for (i in 1:length(vartime)){
trans[i,]= qmsm$estimates[which(qmsm$estimates>0)]
}
trans=as.data.frame(trans)
for (i in 1:length(qmsm$estimates[which(qmsm$estimates>0)]) ){
colnames(trans)[i]<-paste0("Haz_",tr_start_state[i],"_to_",tr_end_state[i])
}
}
if (ci.json=="none" ) {
trans=matrix(nrow=length(vartime), ncol=length(qmsm[which(qmsm>0)]), NA)
for (i in 1:length(vartime)){
trans[i,]= qmsm[which(qmsm>0)]
}
trans=as.data.frame(trans)
for (i in 1:length(qmsm[which(qmsm>0)]) ){
colnames(trans)[i]<-paste0("Haz_",tr_start_state[i],"_to_",tr_end_state[i])
}
}
trans$vartime=vartime
trans
##################### transitions lci and uci #####################################################################
if (ci.json!="none" ) {
trans_lci=matrix(nrow=length(vartime), ncol=length(qmsm$estimates[which(qmsm$L>0)]), NA)
for (i in 1:length(vartime)){
trans_lci[i,]= qmsm$L[which(qmsm$L>0)]
}
trans_lci=as.data.frame(trans_lci)
for (i in 1:length(qmsm$L[which(qmsm$L>0)]) ){
colnames(trans_lci)[i]<- paste0("Haz_",tr_start_state[i],"_to_",tr_end_state[i],"_lci")
}
trans_lci$vartime=vartime
trans_lci
trans_uci=matrix(nrow=length(vartime), ncol=length(qmsm$estimates[which(qmsm$L>0)]), NA)
for (i in 1:length(vartime)){
trans_uci[i,]= qmsm$L[which(qmsm$L>0)]
}
trans_uci=as.data.frame(trans_uci)
for (i in 1:length(qmsm$L[which(qmsm$L>0)]) ){
colnames(trans_uci)[i]<-paste0("Haz_",tr_start_state[i],"_to_",tr_end_state[i],"_uci")
}
trans_uci$vartime=vartime
trans_uci
}
#################################################################################################################
###Los ###
if (totlos==TRUE) {
los_array= array(dim=c(length(vartime),nstates+1,nstates),NA)
los_array_lci= array(dim=c(length(vartime),nstates+1,nstates),NA)
los_array_uci= array(dim=c(length(vartime),nstates+1,nstates),NA)
for (j in states) {
for (i in 1:length(vartime)) {
los.msm <- totlos.msm(msm.model, start=j, fromt=0, tot= vartime[i],covariates=covariates_list[[g]],ci=ci.json,cl=cl.json,
piecewise.times=piecewise.times.json, piecewise.covariates=piecewise.covariates.json,
num.integ=num.integ.json, B=B.json,cores=cores.json)
if (ci.json=="none" ) { los_array[i,,j]=cbind(t(los.msm),vartime=vartime[i] ) }
if (ci.json!="none" ) {
los_array[i,,j]=cbind(t(los.msm[1,]),vartime=vartime[i] )
los_array_lci[i,,j]=cbind(t(round(los.msm[2,],5)),vartime=vartime[i] )
los_array_uci[i,,j]=cbind(t(round(los.msm[3,],5)),vartime=vartime[i] )
}
}
}
loslist=list()
for (k in states) {
loslist[[k]]=los_array[,,k]
loslist[[k]]=as.data.frame(loslist[[k]])
}
if (length(vartime)==1) {
for (k in states) {
loslist[[k]]=as.data.frame(t(loslist[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(loslist[[k]])[j]=paste0("Los_",k,"_to_",j)
}
colnames(loslist[[k]])[nstates+1] ="vartime"
}
loslist
##################### los lci and uci #####################################################################
loslist_lci=list()
loslist_uci=list()
if (ci.json!="none" ) {
###lci
for (k in states) {
loslist_lci[[k]]=los_array_lci[,,k]
loslist_lci[[k]]=as.data.frame(loslist_lci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
loslist_lci[[k]]=as.data.frame(t(loslist_lci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(loslist_lci[[k]])[j]=paste0("Los_",k,"_to_",j,"_lci")
}
colnames(loslist_lci[[k]])[nstates+1] ="vartime"
}
loslist_lci
###uci
for (k in states) {
loslist_uci[[k]]=los_array_uci[,,k]
loslist_uci[[k]]=as.data.frame(loslist_uci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
loslist_uci[[k]]=as.data.frame(t(loslist_uci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(loslist_uci[[k]])[j]=paste0("Los_",k,"_to_",j,"_uci")
}
colnames(loslist_uci[[k]])[nstates+1] ="vartime"
}
loslist_uci
}
}
#################################################################################################################
###Sojourn ###
if (sojourn==TRUE) {
sojlist_=list()
sojlist_lci=list()
sojlist_uci=list()
soj.msm=vector()
if (ci.json!="none") {
soj.msm=sojourn.msm(msm.model,covariates=covariates_list[[g]],ci=ci.json,cl=cl.json, B=B.json)
sojnames= rownames(soj.msm)
sojlist=list()
sojlist[[1]]=cbind.data.frame(t( soj.msm$estimates))
names(sojlist[[1]])[1:(ncol(sojlist[[1]]))]<-paste0("Soj_",sub("State ","",sojnames))
sojlist
sojlist_lci[[1]]=cbind.data.frame(t(soj.msm$L))
sojlist_uci[[1]]=cbind.data.frame(t(soj.msm$U))
names(sojlist_lci[[1]])[1:(ncol(sojlist_lci[[1]]))]<-paste0("Soj_",sub("State ","",sojnames),"_lci")
names(sojlist_uci[[1]])[1:(ncol(sojlist_uci[[1]]))]<-paste0("Soj_",sub("State ","",sojnames),"_uci")
}
else if (ci.json =="none") {
soj.msm=sojourn.msm(msm.model,covariates=covariates_list[[g]],ci=ci.json,cl=cl.json, B=B.json)
sojlist=list()
sojlist[[1]]=cbind.data.frame(t( soj.msm$estimates))
sojnames=c(1:ncol(sojlist[[1]]))
names(sojlist[[1]])[1:(ncol(sojlist[[1]]))]<-paste0("Soj_",sub("State ","",sojnames))
sojlist
}
}
#################################################################################################################
###Next ###
########################### Main estimates ####################################################
if (pnext==TRUE) {
next_array= array(dim=c(1,nstates,nstates),NA)
next.msm <- pnext.msm(msm.model,ci=ci.json,cl=cl.json, B=B.json,cores=cores.json, covariates=covariates_list[[g]])
next_array[1,,]=t(cbind.data.frame(next.msm$estimates ))
nextlist=list()
for (k in states) {
nextlist[[k]]=next_array[,,k]
nextlist[[k]]=as.data.frame(t(nextlist[[k]]))
}
# if (length(vartime)==1) {
# for (k in states) {
# nextlist[[k]]=as.data.frame(t(nextlist[[k]]))
# }
# }
for (k in states) {
for (j in states) {
colnames(nextlist[[k]])[j]=paste0("Next_",k,"_to_",j)
}
}
nextlist
########################### next uci , lci ####################################################
nextlist_lci=list()
nextlist_uci=list()
if (ci.json!="none" ) {
###lci
next_array_lci= array(dim=c(1,nstates,nstates),NA)
next_array_lci[1,,]=t(cbind.data.frame(next.msm$L ))
for (k in states) {
nextlist_lci[[k]]=next_array_lci[,,k]
nextlist_lci[[k]]=as.data.frame(t(nextlist_lci[[k]]))
}
# if (length(vartime)==1) {
# for (k in states) {
# nextlist_lci[[k]]=as.data.frame(t(nextlist_lci[[k]]))
# }
# }
for (k in states) {
for (j in states) {
colnames(nextlist_lci[[k]])[j]=paste0("Next_",k,"_to_",j,"_lci")
}
}
nextlist_lci
### uci
next_array_uci= array(dim=c(1,nstates,nstates),NA)
next_array_uci[1,,]=t(cbind.data.frame(next.msm$U ))
for (k in states) {
nextlist_uci[[k]]=next_array_uci[,,k]
nextlist_uci[[k]]=as.data.frame(t(nextlist_uci[[k]]))
}
# if (length(vartime)==1) {
# for (k in states) {
# nextlist_uci[[k]]=as.data.frame(t(nextlist_uci[[k]]))
# }
# }
for (k in states) {
for (j in states) {
colnames(nextlist_uci[[k]])[j]=paste0("Next_",k,"_to_",j,"_uci")
}
}
nextlist_uci
}
}
###############################################################################
##First
if (efpt==TRUE) {
first.msm=vector()
first.msm= efpt.msm(msm.model, start="all", tostate=states,
ci=ci.json,cl=cl.json, B=B.json,cores=cores.json, covariates = covariates_list[[g]])
firstlist=list()
if (ci.json=="none" ) {
firstlist[[1]]=cbind.data.frame(t( first.msm))
}
if (ci.json!="none" ) {
firstlist[[1]]=cbind.data.frame(t( first.msm[1,]))
}
for (i in 1:nstates) {
names( firstlist[[1]])[i]<-paste0("First_",i)
}
firstlist_lci=list()
firstlist_uci=list()
if (ci.json!="none" ) {
firstlist_lci[[1]]=cbind.data.frame(t(first.msm[2,]))
firstlist_uci[[1]]=cbind.data.frame(t(first.msm[3,]))
for (i in 1:nstates) {
names( firstlist_lci[[1]])[i]<-paste0("First_",i,"_lci")
names( firstlist_uci[[1]])[i]<-paste0("First_",i,"_uci")
}
}
}
##########################################################################################
##Number
if (envisits==TRUE) {
############Main estimates ########################
number_array= array(dim=c(length(vartime),nstates+1,nstates),NA)
number_array_lci= array(dim=c(length(vartime),nstates+1,nstates),NA)
number_array_uci= array(dim=c(length(vartime),nstates+1,nstates),NA)
for (j in states) {
for (i in 1:length(vartime)) {
number.msm <- envisits.msm(msm.model, start=j, fromt=0, tot= vartime[i],covariates=covariates_list[[g]],
piecewise.times=piecewise.times.json, piecewise.covariates=piecewise.covariates.json,
num.integ=num.integ.json,
ci=ci.json, cl=cl.json, B=B.json,
cores=cores.json)
if (ci.json=="none" ) { number_array[i,,j]=cbind(t(number.msm),vartime=vartime[i] ) }
if (ci.json!="none" ) {
number_array[i,,j]=cbind(t(round(number.msm[1,],5)),vartime=vartime[i] )
number_array_lci[i,,j]=cbind(t(round(number.msm[2,],5)),vartime=vartime[i] )
number_array_uci[i,,j]=cbind(t(round(number.msm[3,],5)),vartime=vartime[i] )
}
}
}
numberlist=list()
for (k in states) {
numberlist[[k]]=number_array[,,k]
numberlist[[k]]=as.data.frame(numberlist[[k]])
}
if (length(vartime)==1) {
for (k in states) {
numberlist[[k]]=as.data.frame(t(numberlist[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(numberlist[[k]])[j]=paste0("Number_",k,"_to_",j)
}
colnames(numberlist[[k]])[nstates+1] ="vartime"
}
numberlist
numberlist_lci=list()
numberlist_uci=list()
if (ci.json!="none" ) {
############ lci estimates ########################
for (k in states) {
numberlist_lci[[k]]=number_array_lci[,,k]
numberlist_lci[[k]]=as.data.frame(numberlist_lci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
numberlist_lci[[k]]=as.data.frame(t(numberlist_lci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(numberlist_lci[[k]])[j]=paste0("Number_",k,"_to_",j,"_lci")
}
colnames(numberlist_lci[[k]])[nstates+1] ="vartime"
}
numberlist_lci
############ uci estimates ########################
for (k in states) {
numberlist_uci[[k]]=number_array_uci[,,k]
numberlist_uci[[k]]=as.data.frame(numberlist_uci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
numberlist_uci[[k]]=as.data.frame(t(numberlist_uci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(numberlist_uci[[k]])[j]=paste0("Number_",k,"_to_",j,"_uci")
}
colnames(numberlist_uci[[k]])[nstates+1] ="vartime"
}
numberlist_uci
}
}
###########################################################################################################
##Visit
if (visit==TRUE) {
visit.msm=vector()
######## What should we put in start? 1 or all states
visit_array= array(dim=c(length(vartime),nstates+1,nstates),NA)
visit_array_lci= array(dim=c(length(vartime),nstates+1,nstates),NA)
visit_array_uci= array(dim=c(length(vartime),nstates+1,nstates),NA)
for (i in 1:length(vartime)) {
visit.msm <- ppass.msm(x=msm.model, qmatrix=NULL, tot=vartime[i], start="all", covariates = covariates_list[[g]],
piecewise.times=piecewise.times.json, piecewise.covariates=piecewise.covariates.json,
ci=ci.json, cl=cl.json, B=B.json,
cores=cores.json)
if (ci.json=="none" ) {visit_array[i,,]=t(cbind(visit.msm,vartime=vartime[i] )) }
if (ci.json!="none" ) {
visit_array[i,,]=t(cbind(visit.msm$estimates,vartime=vartime[i] ))
visit_array_lci[i,,]=t(cbind(visit.msm$L,vartime=vartime[i] ))
visit_array_uci[i,,]=t(cbind(visit.msm$U,vartime=vartime[i] ))
}
}
####################### Main estimates ########################
vlist=list()
for (k in states) {
vlist[[k]]=visit_array[,,k]
vlist[[k]]=as.data.frame(vlist[[k]])
}
if (length(vartime)==1) {
for (k in states) {
vlist[[k]]=as.data.frame(t(vlist[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(vlist[[k]])[j]=paste0("Visit_",k,"_to_",j)
}
colnames(vlist[[k]])[nstates+1] ="vartime"
}
vlist
vlist_lci=list()
vlist_uci=list()
if (ci.json!="none" ) {
####################### Lci estimates ########################
for (k in states) {
vlist_lci[[k]]=visit_array_lci[,,k]
vlist_lci[[k]]=as.data.frame(vlist_lci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
vlist_lci[[k]]=as.data.frame(t(vlist_lci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(vlist_lci[[k]])[j]=paste0("Visit_",k,"_to_",j,"_lci")
}
colnames(vlist_lci[[k]])[nstates+1] ="vartime"
}
vlist_lci
####################### Uci estimates ########################
for (k in states) {
vlist_uci[[k]]=visit_array_uci[,,k]
vlist_uci[[k]]=as.data.frame(vlist_uci[[k]])
}
if (length(vartime)==1) {
for (k in states) {
vlist_uci[[k]]=as.data.frame(t(vlist_uci[[k]]))
}
}
for (k in states) {
for (j in states) {
colnames(vlist_uci[[k]])[j]=paste0("Visit_",k,"_to_",j,"_uci")
}
colnames(vlist_uci[[k]])[nstates+1] ="vartime"
}
vlist_uci
}
}
###########################################################################################
# ###Ratio of tr. intensities
# mat
#
# p=1
# qratio=matrix(nrow=1, ncol=length(which(mat>0))^2,NA)
# names.qratio=vector()
#
# for (i in 1:nstates) {
# for (j in 1:nstates) {
# for (k in 1:nstates) {
# for (f in 1:nstates) {
#
# if (mat[k,f]==0 | mat[i,j]==0 | k==f | i==j) {next}
# else if (mat[k,f]!=0 & mat[i,j]!=0 & k!=f & i!=j ) {
# qratio[1,p]=qratio.msm(msm.model, ind1=c(i,j), ind2=c(k,f), covariates = covariates_list[[g]])[1]
# names.qratio[p]=paste0("Qratio_t1_",i,".",j,"_t2_",k,".",f)
# p=p+1
# }
# }
# }
# }
# }
#
# qratio=as.data.frame(qratio)
# names(qratio)=names.qratio
#
# qratiolist=list()
# qratiolist[[1]]=cbind.data.frame(qratio,cov_factor=rep(paste0(names(covariates_list[[g]]),covariates_list[[g]]),1))
##############################################################################################
pred[[g]]= list(probs=plist,probs_lci=plist_lci,probs_uci=plist_uci,
trans=trans, trans_lci=trans_lci,trans_uci=trans_uci,
los=loslist,los_lci=loslist_lci,los_uci=loslist_uci,
soj=sojlist,soj_lci=sojlist_lci,soj_uci=sojlist_uci,
nextv=nextlist,nextv_lci=nextlist_lci,nextv_uci=nextlist_uci,
first=firstlist,first_lci=firstlist_lci,first_uci=firstlist_uci,
number=numberlist,number_lci=numberlist_lci,number_uci=numberlist_uci,
visit=vlist,visit_lci=vlist_lci, visit_uci=vlist_uci
)
}
}
########################################################################################################
######## Declare the list elements even if they stay empty ####
pjson_lci=list()
pjson_uci=list()
hjson_lci=list()
hjson_uci=list()
losjson=list()
losjson_lci=list()
losjson_uci=list()
sojjson=list()
sojjson_lci=list()
sojjson_uci=list()
nextjson=list()
nextjson_lci=list()
nextjson_uci=list()
firstjson=list()
firstjson_lci=list()
firstjson_uci=list()
numberjson=list()
numberjson_lci=list()
numberjson_uci=list()
vjson=list()
vjson_lci=list()
vjson_uci=list()
########################################################################################################
#### Probabilities
########################################################################################################
##### Main estimates
pjson=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
pjson[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$probs[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$probs[[k]])[i]), pjson[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplist=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplist[1,]=as.vector(pred[[1]]$probs[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplist[j,]=as.vector(pred[[j]]$probs[[k]][,i])
}
pjson[[p]]=tmplist
assign(paste0("forMSM",names(pred[[1]]$probs[[k]])[i]), pjson[[p]])
p=p+1
}
}
}
pvector=vector()
pvector=ls(pattern = "forMSMP_._to_.$")
pvector=sub("forMSM","",ls(pattern = "^forMSMP_._to_.$") )
pvector=stringi::stri_sort(pvector, numeric = TRUE)
names(pjson)=pvector[1:(p-1)]
pjson
##### Lci possibilities estimates
if (ci.json!="none" ) {
pjson_lci=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
pjson_lci[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$probs_lci[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$probs_lci[[k]])[i]), pjson_lci[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplist=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplist[1,]=as.vector(pred[[1]]$probs_lci[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplist[j,]=as.vector(pred[[j]]$probs_lci[[k]][,i])
}
pjson_lci[[p]]=tmplist
assign(paste0("forMSM",names(pred[[1]]$probs_lci[[k]])[i]), pjson_lci[[p]])
p=p+1
}
}
}
pvector_lci=vector()
pvector_lci=ls(pattern = "^forMSMP.*lci$")
pvector_lci=sub("forMSM","",ls(pattern = "^forMSMP.*lci$") )
pvector_lci=stringi::stri_sort(pvector_lci, numeric = TRUE)
names(pjson_lci)=pvector_lci[1:(p-1)]
##### Uci possibilities estimates
pjson_uci=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
pjson_uci[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$probs_uci[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$probs_uci[[k]])[i]), pjson_uci[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplist=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplist[1,]=as.vector(pred[[1]]$probs_uci[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplist[j,]=as.vector(pred[[j]]$probs_uci[[k]][,i])
}
pjson_uci[[p]]=tmplist
assign(paste0("forMSM",names(pred[[1]]$probs_uci[[k]])[i]), pjson_uci[[p]])
p=p+1
}
}
}
pvector_uci=vector()
pvector_uci=ls(pattern = "^forMSMP.*uci$")
pvector_uci=sub("forMSM","",ls(pattern = "^forMSMP.*uci$") )
pvector_uci=stringi::stri_sort(pvector_uci, numeric = TRUE)
names(pjson_uci)=pvector_uci[1:(p-1)]
}
#################################################################################################################
### Transitions main estimates ###
hjson=list()
if (length(covariates_list)<=1) {
for (k in 1:ntransitions) {
tmplisth=matrix(nrow =1 , ncol= length(vartime))
tmplisth[1,]=as.vector(pred[[1]]$trans[,k])
hjson[[1]]=tmplisth
assign(paste0("forMSM",names(pred[[1]]$trans)[i]), hjson[[1]])
}
}
if (length(covariates_list)>1) {
for (k in 1:ntransitions) {
tmplisth=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplisth[1,]=as.vector(pred[[1]]$trans[,k])
for (j in 2:length(covariates_list)) {
tmplisth[j,]=as.vector(pred[[j]]$trans[,k])
}
hjson[[k]]=tmplisth
assign(paste0("forMSM",names(pred[[1]]$trans)[k]), hjson[[k]])
}
}
hvector=vector()
hvector=ls(pattern = "^forMSMHaz_._to_.$")
hvector=sub("forMSM","",ls(pattern = "^forMSMHaz_._to_.$") )
hvector=stringi::stri_sort(hvector, numeric = TRUE)
names(hjson)=hvector
if (ci.json!="none" ) {
### Transitions lci estimates ###
hjson_lci=list()
if (length(covariates_list)<=1) {
for (k in 1:ntransitions) {
tmplisth=matrix(nrow =1 , ncol= length(vartime))
tmplisth[1,]=as.vector(pred[[1]]$trans_lci[,k])
hjson_lci[[1]]=tmplisth
assign(paste0("forMSM",names(pred[[1]]$trans_lci)[i]), hjson_lci[[1]])
}
}
if (length(covariates_list)>1) {
for (k in 1:ntransitions) {
tmplisth=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplisth[1,]=as.vector(pred[[1]]$trans_lci[,k])
for (j in 2:length(covariates_list)) {
tmplisth[j,]=as.vector(pred[[j]]$trans_lci[,k])
}
hjson_lci[[k]]=tmplisth
assign(paste0("forMSM",names(pred[[1]]$trans_lci)[k]), hjson_lci[[k]])
}
}
hvector_lci=vector()
hvector_lci=ls(pattern = "^forMSMHaz_._to_._lci$")
hvector_lci=sub("forMSM","",ls(pattern = "^forMSMHaz_._to_._lci$") )
hvector_lci=stringi::stri_sort(hvector_lci, numeric = TRUE)
names(hjson_lci)=hvector_lci
### Transitions uci estimates ###
hjson_uci=list()
if (length(covariates_list)<=1) {
for (k in 1:ntransitions) {
tmplisth=matrix(nrow =1 , ncol= length(vartime))
tmplisth[1,]=as.vector(pred[[1]]$trans_uci[,k])
hjson_uci[[1]]=tmplisth
assign(paste0("forMSM",names(pred[[1]]$trans_uci)[i]), hjson_uci[[1]])
}
}
if (length(covariates_list)>1) {
for (k in 1:ntransitions) {
tmplisth=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplisth[1,]=as.vector(pred[[1]]$trans_uci[,k])
for (j in 2:length(covariates_list)) {
tmplisth[j,]=as.vector(pred[[j]]$trans_uci[,k])
}
hjson_uci[[k]]=tmplisth
assign(paste0("forMSM",names(pred[[1]]$trans_uci)[k]), hjson_uci[[k]])
}
}
hvector_uci=vector()
hvector_uci=ls(pattern = "^forMSMHaz_._to_._uci$")
hvector_uci=sub("forMSM","",ls(pattern = "^forMSMHaz_._to_._uci$") )
hvector_uci=stringi::stri_sort(hvector_uci, numeric = TRUE)
names(hjson_uci)=hvector_uci
}
##########################################################################
### Los ################################################################
##############################################################################
if (totlos==TRUE) {
### Los Main estimates #####
losjson=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
losjson[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$los[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$los[[k]])[i]), losjson[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplistlos=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplistlos[1,]=as.vector(pred[[1]]$los[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplistlos[j,]=as.vector(pred[[j]]$los[[k]][,i])
}
losjson[[p]]=tmplistlos
assign(paste0("forMSM",names(pred[[1]]$los[[k]])[i]), losjson[[p]])
p=p+1
}
}
}
losvector=vector()
losvector=ls(pattern = "forMSMLos_._to_.$")
losvector=sub("forMSM","",ls(pattern = "^forMSMLos_._to_.$") )
losvector=stringi::stri_sort(losvector, numeric = TRUE)
names(losjson)=losvector
if (ci.json!="none" ) {
### Los lci estimates #####
losjson_lci=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
losjson_lci[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$los_lci[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$los_lci[[k]])[i]), losjson_lci[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplistlos=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplistlos[1,]=as.vector(pred[[1]]$los_lci[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplistlos[j,]=as.vector(pred[[j]]$los_lci[[k]][,i])
}
losjson_lci[[p]]=tmplistlos
assign(paste0("forMSM",names(pred[[1]]$los_lci[[k]])[i]), losjson_lci[[p]])
p=p+1
}
}
}
losvector_lci=vector()
losvector_lci=ls(pattern = "^forMSMLos.*lci$")
losvector_lci=sub("forMSM","",ls(pattern = "^forMSMLos.*lci$") )
losvector_lci=stringi::stri_sort(losvector_lci, numeric = TRUE)
names(losjson_lci)=losvector_lci
### Los uci estimates #####
losjson_uci=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
losjson_uci[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$los_uci[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$los_uci[[k]])[i]), losjson_uci[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplistlos=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplistlos[1,]=as.vector(pred[[1]]$los_uci[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplistlos[j,]=as.vector(pred[[j]]$los_uci[[k]][,i])
}
losjson_uci[[p]]=tmplistlos
assign(paste0("forMSM",names(pred[[1]]$los_uci[[k]])[i]), losjson_uci[[p]])
p=p+1
}
}
}
losvector_uci=vector()
losvector_uci=ls(pattern = "^forMSMLos.*uci$")
losvector_uci=sub("forMSM","",ls(pattern = "^forMSMLos.*uci$") )
losvector_uci=stringi::stri_sort(losvector_uci, numeric = TRUE)
names(losjson_uci)=losvector_uci
}
}
##########################################################################
### Sojourn ################################################################
##############################################################################
if ( sojourn==TRUE) {
### Sojourn Main estimates #####
sojjson=list()
if (length(covariates_list)<=1) {
p=1
for (k in na_states) {
tmplistsoj=matrix(nrow =1 , ncol= 1)
tmplistsoj[1,1]=as.matrix(pred[[1]]$soj[[1]][1])
sojjson[[1]]=tmplistsoj
assign(paste0("forMSM",names(pred[[1]]$soj)[i]), tmplistsoj[[1]])
p=p+1
}
}
if (length(covariates_list)>1) {
p=1
for (k in na_states) {
tmplistsoj=matrix(nrow =length(covariates_list) , ncol= 1)
for (j in 1:length(covariates_list)) {
tmplistsoj[j,]=as.matrix(pred[[j]]$soj[[1]][k])
}
sojjson[[p]]=tmplistsoj
assign(paste0("forMSM",names(pred[[1]]$soj[[1]])[k]), sojjson[[p]])
p=p+1
}
}
sojvector=vector()
sojvector=ls(pattern = "forMSMSoj_.$")
sojvector=sub("forMSM","",ls(pattern = "^forMSMSoj_.$") )
sojvector=stringi::stri_sort(sojvector, numeric = TRUE)
names(sojjson)=sojvector
if (ci.json!="none" ) {
### Sojourn lci estimates #####
sojjson_lci=list()
if (length(covariates_list)<=1) {
p=1
for (k in na_states) {
tmplistsoj=matrix(nrow =1 , ncol= 1)
tmplistsoj[1,1]=as.matrix(pred[[1]]$soj_lci[[1]][1])
sojjson_lci[[1]]=tmplistsoj
assign(paste0("forMSM",names(pred[[1]]$soj_lci)[i]), tmplistsoj[[1]])
p=p+1
}
}
if (length(covariates_list)>1) {
p=1
for (k in na_states) {
tmplistsoj=matrix(nrow =length(covariates_list) , ncol= 1)
for (j in 1:length(covariates_list)) {
tmplistsoj[j,]=as.matrix(pred[[j]]$soj_lci[[1]][k])
}
sojjson_lci[[p]]=tmplistsoj
assign(paste0("forMSM",names(pred[[1]]$soj_lci[[1]])[k]), sojjson_lci[[p]])
p=p+1
}
}
sojvector_lci=vector()
sojvector_lci=ls(pattern = "^forMSMSoj.*lci$")
sojvector_lci=sub("forMSM","",ls(pattern = "^forMSMSoj.*lci$") )
sojvector_lci=stringi::stri_sort(sojvector_lci, numeric = TRUE)
names(sojjson_lci)=sojvector_lci
### Sojourn uci estimates #####
sojjson_uci=list()
if (length(covariates_list)<=1) {
p=1
for (k in na_states) {
tmplistsoj=matrix(nrow =1 , ncol= 1)
tmplistsoj[1,1]=as.matrix(pred[[1]]$soj_uci[[1]][1])
sojjson_uci[[1]]=tmplistsoj
assign(paste0("forMSM",names(pred[[1]]$soj_uci)[i]), tmplistsoj[[1]])
p=p+1
}
}
if (length(covariates_list)>1) {
p=1
for (k in na_states) {
tmplistsoj=matrix(nrow =length(covariates_list) , ncol= 1)
for (j in 1:length(covariates_list)) {
tmplistsoj[j,]=as.matrix(pred[[j]]$soj_uci[[1]][k])
}
sojjson_uci[[p]]=tmplistsoj
assign(paste0("forMSM",names(pred[[1]]$soj_uci[[1]])[k]), sojjson_uci[[p]])
p=p+1
}
}
sojvector_uci=vector()
sojvector_uci=ls(pattern = "^forMSMSoj.*uci$")
sojvector_uci=sub("forMSM","",ls(pattern = "^forMSMSoj.*uci$") )
sojvector_uci=stringi::stri_sort(sojvector_uci, numeric = TRUE)
names(sojjson_uci)=sojvector_uci
}
}
##########################################################################
### Next ################################################################
##############################################################################
if (pnext==TRUE) {
### Next Main estimates #####
nextjson=list()
if (length(covariates_list)<=1) {
for (k in 1:nstates) {
tmplistnext=matrix(nrow =1 , ncol= 1)
tmplistnext[1,1]=as.matrix(pred[[1]]$nextv[[1]][1])
nextjson[[1]]=tmplistnext
assign(paste0("forMSM",names(pred[[1]]$nextv)[i]), tmplistnext[[1]])
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
tmplistnext=matrix(nrow =length(covariates_list) , ncol= 1)
for (j in 1:length(covariates_list)) {
tmplistnext[j,]=as.matrix(pred[[j]]$nextv[[1]][k])
}
nextjson[[p]]=tmplistnext
assign(paste0("forMSM",names(pred[[1]]$nextv[[1]])[k]), nextjson[[p]])
p=p+1
}
}
nextvector=vector()
nextvector=ls(pattern = "forMSMNext_._to_.$")
nextvector=sub("forMSM","",ls(pattern = "^forMSMNext_._to_.$") )
nextvector=stringi::stri_sort(nextvector, numeric = TRUE)
names(nextjson)=nextvector
if (ci.json!="none" ) {
####### Next lci estimates ####################################
nextjson_lci=list()
if (length(covariates_list)<=1) {
for (k in 1:nstates) {
tmplistnext=matrix(nrow =1 , ncol= 1)
tmplistnext[1,1]=as.matrix(pred[[1]]$nextv_lci[[1]][1])
nextjson_lci[[1]]=tmplistnext
assign(paste0("forMSM",names(pred[[1]]$nextv_lci)[i]), tmplistnext[[1]])
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
tmplistnext=matrix(nrow =length(covariates_list) , ncol= 1)
for (j in 1:length(covariates_list)) {
tmplistnext[j,]=as.matrix(pred[[j]]$nextv_lci[[1]][k])
}
nextjson_lci[[p]]=tmplistnext
assign(paste0("forMSM",names(pred[[1]]$nextv_lci[[1]])[k]), nextjson_lci[[p]])
p=p+1
}
}
nextvector_lci=vector()
nextvector_lci=ls(pattern = "^forMSMNext.*lci$")
nextvector_lci=sub("forMSM","",ls(pattern = "^forMSMNext.*lci$") )
nextvector_lci=stringi::stri_sort(nextvector_lci, numeric = TRUE)
names(nextjson_lci)=nextvector_lci
####### Next uci estimates ####################################
nextjson_uci=list()
if (length(covariates_list)<=1) {
for (k in 1:nstates) {
tmplistnext=matrix(nrow =1 , ncol= 1)
tmplistnext[1,1]=as.matrix(pred[[1]]$nextv_uci[[1]][1])
nextjson_uci[[1]]=tmplistnext
assign(paste0("forMSM",names(pred[[1]]$nextv_uci)[i]), tmplistnext[[1]])
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
tmplistnext=matrix(nrow =length(covariates_list) , ncol= 1)
for (j in 1:length(covariates_list)) {
tmplistnext[j,]=as.matrix(pred[[j]]$nextv_uci[[1]][k])
}
nextjson_uci[[p]]=tmplistnext
assign(paste0("forMSM",names(pred[[1]]$nextv_uci[[1]])[k]), nextjson_uci[[p]])
p=p+1
}
}
nextvector_uci=vector()
nextvector_uci=ls(pattern = "^forMSMNext.*uci$")
nextvector_uci=sub("forMSM","",ls(pattern = "^forMSMNext.*uci$") )
nextvector_uci=stringi::stri_sort(nextvector_uci, numeric = TRUE)
names(nextjson_uci)=nextvector_uci
}
}
###################################################################################################
##########################################################################
### First ################################################################
##############################################################################
if (efpt==TRUE) {
### First Main estimates #####
firstjson=list()
if (length(covariates_list)<=1) {
for (k in 1:3) {
tmplistfirst=matrix(nrow =1 , ncol= 1)
tmplistfirst[1,1]=as.matrix(pred[[1]]$first[[1]][1])
firstjson[[1]]=tmplistfirst
assign(paste0("forMSM",names(pred[[1]]$first)[i]), tmplistfirst[[1]])
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:3) {
tmplistfirst=matrix(nrow =length(covariates_list) , ncol= 1)
for (j in 1:length(covariates_list)) {
tmplistfirst[j,]=as.matrix(pred[[j]]$first[[1]][k])
}
firstjson[[p]]=tmplistfirst
assign(paste0("forMSM",names(pred[[1]]$first[[1]])[k]), firstjson[[p]])
p=p+1
}
}
firstvector=vector()
firstvector=ls(pattern = "forMSMFirst_.$")
firstvector=sub("forMSM","",ls(pattern = "^forMSMFirst") )
firstvector=stringi::stri_sort(firstvector, numeric = TRUE)
names(firstjson)=firstvector
if (ci.json!="none" ) {
####### First lci estimates ####################################
firstjson_lci=list()
if (length(covariates_list)<=1) {
for (k in 1:nstates) {
tmplistfirst=matrix(nrow =1, ncol= 1)
tmplistfirst[1,1]=as.matrix(pred[[1]]$first_lci[[1]][1])
firstjson_lci[[1]]=tmplistfirst
assign(paste0("forMSM",names(pred[[1]]$first_lci)[i]), tmplistfirst[[1]])
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
tmplistfirst=matrix(nrow =length(covariates_list) , ncol= 1)
for (j in 1:length(covariates_list)) {
tmplistfirst[j,]=as.matrix(pred[[j]]$first_lci[[1]][k])
}
firstjson_lci[[p]]=tmplistfirst
assign(paste0("forMSM",names(pred[[1]]$first_lci[[1]])[k]), firstjson_lci[[p]])
p=p+1
}
}
firstvector_lci=vector()
firstvector_lci=ls(pattern = "^forMSMFirst.*lci$")
firstvector_lci=sub("forMSM","",ls(pattern ="^forMSMFirst.*lci$" ) )
firstvector_lci=stringi::stri_sort(firstvector_lci, numeric = TRUE)
names(firstjson_lci)=firstvector_lci
####### First uci estimates ####################################
firstjson_uci=list()
if (length(covariates_list)<=1) {
for (k in 1:nstates) {
tmplistfirst=matrix(nrow =1 , ncol= 1)
tmplistfirst[1,1]=as.matrix(pred[[1]]$first_uci[[1]][1])
firstjson_uci[[1]]=tmplistfirst
assign(paste0("forMSM",names(pred[[1]]$first_uci)[i]), tmplistfirst[[1]])
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
tmplistfirst=matrix(nrow =length(covariates_list) , ncol= 1)
for (j in 1:length(covariates_list)) {
tmplistfirst[j,]=as.matrix(pred[[j]]$first_uci[[1]][k])
}
firstjson_uci[[p]]=tmplistfirst
assign(paste0("forMSM",names(pred[[1]]$first_uci[[1]])[k]), firstjson_uci[[p]])
p=p+1
}
}
firstvector_uci=vector()
firstvector_uci=ls(pattern = "^forMSMFirst.*uci$")
firstvector_uci=sub("forMSM","",ls(pattern ="^forMSMFirst.*uci$" ) )
firstvector_uci=stringi::stri_sort(firstvector_uci, numeric = TRUE)
names(firstjson_uci)=firstvector_uci
}
}
################################################################################################
##########################################################################
### Number ################################################################
##############################################################################
if (envisits==TRUE) {
### Number Main estimates #####
numberjson=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
numberjson[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$number[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$number[[k]])[i]), numberjson[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplistnumber=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplistnumber[1,]=as.vector(pred[[1]]$number[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplistnumber[j,]=as.vector(pred[[j]]$number[[k]][,i])
}
numberjson[[p]]=tmplistnumber
assign(paste0("forMSM",names(pred[[1]]$number[[k]])[i]), numberjson[[p]])
p=p+1
}
}
}
numbervector=vector()
numbervector=ls(pattern = "forMSMNumber_._to_.$")
numbervector=sub("forMSM","",ls(pattern = "^forMSMNumber_._to_.$") )
numbervector=stringi::stri_sort(numbervector, numeric = TRUE)
names(numberjson)=numbervector
if (ci.json!="none" ) {
### Number lci estimates #####
numberjson_lci=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
numberjson_lci[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$number_lci[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$number_lci[[k]])[i]), numberjson_lci[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplistnumber=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplistnumber[1,]=as.vector(pred[[1]]$number_lci[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplistnumber[j,]=as.vector(pred[[j]]$number_lci[[k]][,i])
}
numberjson_lci[[p]]=tmplistnumber
assign(paste0("forMSM",names(pred[[1]]$number_lci[[k]])[i]), numberjson_lci[[p]])
p=p+1
}
}
}
numbervector_lci=vector()
numbervector_lci=ls(pattern = "^forMSMNumber.*lci$")
numbervector_lci=sub("forMSM","",ls(pattern = "^forMSMNumber.*lci$") )
numbervector_lci=stringi::stri_sort(numbervector_lci, numeric = TRUE)
names(numberjson_lci)=numbervector_lci
### Number uci estimates #####
numberjson_uci=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
numberjson_uci[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$number_uci[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$number_uci[[k]])[i]), numberjson_uci[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplistnumber=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplistnumber[1,]=as.vector(pred[[1]]$number_uci[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplistnumber[j,]=as.vector(pred[[j]]$number_uci[[k]][,i])
}
numberjson_uci[[p]]=tmplistnumber
assign(paste0("forMSM",names(pred[[1]]$number_uci[[k]])[i]), numberjson_uci[[p]])
p=p+1
}
}
}
numbervector_uci=vector()
numbervector_uci=ls(pattern = "^forMSMNumber.*uci$")
numbervector_uci=sub("forMSM","",ls(pattern = "^forMSMNumber.*uci$") )
numbervector_uci=stringi::stri_sort(numbervector_uci, numeric = TRUE)
names(numberjson_uci)=numbervector_uci
}
}
######################################################################################################
##########################################################################
### Visit ################################################################
##############################################################################
if (visit==TRUE) {
### Visit Main estimates #####
vjson=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
vjson[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$visit[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$visit[[k]])[i]), vjson[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplistvisit=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplistvisit[1,]=as.vector(pred[[1]]$visit[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplistvisit[j,]=as.vector(pred[[j]]$visit[[k]][,i])
}
vjson[[p]]=tmplistvisit
assign(paste0("forMSM",names(pred[[1]]$visit[[k]])[i]), vjson[[p]])
p=p+1
}
}
}
visvector=vector()
visvector=ls(pattern = "forMSMVisit_._to_.$")
visvector=sub("forMSM","",ls(pattern = "^forMSMVisit_._to_.$") )
visvector=stringi::stri_sort(visvector, numeric = TRUE)
names(vjson)=visvector
if (ci.json!="none" ) {
### Visit lci estimates #####
vjson_lci=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
vjson_lci[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$visit_lci[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$visit_lci[[k]])[i]), vjson_lci[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplistvisit=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplistvisit[1,]=as.vector(pred[[1]]$visit_lci[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplistvisit[j,]=as.vector(pred[[j]]$visit_lci[[k]][,i])
}
vjson_lci[[p]]=tmplistvisit
assign(paste0("forMSM",names(pred[[1]]$visit_lci[[k]])[i]), vjson_lci[[p]])
p=p+1
}
}
}
visvector_lci=vector()
visvector_lci=ls(pattern = "^forMSMVisit.*lci$")
visvector_lci=sub("forMSM","",ls(pattern ="^forMSMVisit.*lci$") )
visvector_lci=stringi::stri_sort(visvector_lci, numeric = TRUE)
names(vjson_lci)=visvector_lci[1:(p-1)]
### Visit uci estimates #####
vjson_uci=list()
if (length(covariates_list)<=1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
vjson_uci[[p]]= as.data.frame(rbind(as.vector(pred[[1]]$visit_uci[[k]][,i])))
assign(paste0("forMSM",names(pred[[1]]$visit_uci[[k]])[i]), vjson_uci[[p]])
p=p+1
}
}
}
if (length(covariates_list)>1) {
p=1
for (k in 1:nstates) {
for (i in 1:nstates) {
tmplistvisit=matrix(nrow =length(covariates_list) , ncol= length(vartime))
tmplistvisit[1,]=as.vector(pred[[1]]$visit_uci[[k]][,i])
for (j in 2:length(covariates_list)) {
tmplistvisit[j,]=as.vector(pred[[j]]$visit_uci[[k]][,i])
}
vjson_uci[[p]]=tmplistvisit
assign(paste0("forMSM",names(pred[[1]]$visit_uci[[k]])[i]), vjson_uci[[p]])
p=p+1
}
}
}
visvector_uci=vector()
visvector_uci=ls(pattern = "^forMSMVisit.*uci$")
visvector_uci=sub("forMSM","",ls(pattern ="^forMSMVisit.*uci$") )
visvector_uci=stringi::stri_sort(visvector_uci, numeric = TRUE)
names(vjson_uci)=visvector_uci[1:(p-1)]
}
}
######################################################################################################
##########################################################################################
### vartime###
timejson=list()
#timematrix=matrix(nrow=1, ncol=length(vartime), NA)
#timematrix[1,]=vartime
timejson[[1]]=vartime
######################################################################################################
### At list ###
atlistjson=list()
if (length(covariates_list)>=1) {
atlistmatrix=matrix(nrow=1, ncol=length(covariates_list), NA)
for (j in 1:length(covariates_list) ) {
name_paste=list()
for (i in 1:length(covariates_list[[j]]) ) {
name_paste[[i]]= paste0(names(covariates_list[[j]])[i]," ", as.character(covariates_list[[j]][i]) )
}
atlistmatrix[1,j]= paste(unlist(name_paste,recursive=FALSE), collapse = ' ')
}
}
if (length(covariates_list)==0) {
atlistmatrix=matrix(nrow=1, ncol=1, NA)
name_paste= "all"
atlistmatrix[1,1]= name_paste
}
atlistjson[[1]]=as.vector(atlistmatrix)
######################################################################################################
final_list=list(timevar=timejson, Nats=length(covariates_list) ,
atlist=atlistjson, tmat=tmatlist, intermediate_states=intermediate_list, Ntransitions= ntransitions,
pjson,pjson_lci,pjson_uci,
hjson,hjson_lci,hjson_uci,
losjson,losjson_lci,losjson_uci,
sojjson,sojjson_lci,sojjson_uci,
nextjson,nextjson_lci,nextjson_uci,
firstjson, firstjson_lci, firstjson_uci,
numberjson,numberjson_lci,numberjson_uci,
vjson,vjson_lci,vjson_uci
)
final_list$timevar=as.vector(final_list$timevar)
# final_list$atlistjson=as.vector(final_list$atlistjson[[1]])
final_unlist= unlist(final_list, recursive=FALSE)
exportJson <- RJSONIO::toJSON(final_unlist, pretty = TRUE,force = TRUE, flatten=TRUE, na='string')
#exportJson
wd=getwd()
if (is.null(jsonpath) ) {
write(exportJson , paste0(wd ,"/", name ) )
}
if (!is.null(jsonpath)) {
if (nchar(jsonpath)==0) {
write(exportJson , paste0(wd ,"/", name ) )
}
if (nchar(jsonpath)!=0) {
write(exportJson, paste0(jsonpath ,"/", name ) )
}
}
rm(list=ls(pattern="^forMSM"))
final_list
}
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.