Description Usage Arguments Value See Also Examples
Copula models for count response data, including discretized MVN diagnostics from bivariate margins
1 2 3 4 5 6 7 8 9 10 11 12 | latentBVNnllk1(rho,param,ucdf,xdat1,xdat2,y1,y2)
latentBVNnllk2(rho,par1,par2,cdf1,cdf2,xdat1,xdat2,y1,y2)
ieenllk(param,upmf,ydat,xdat,LB,UB) # independent estimating equations
MVNlatent1(ydat,xdat,nrep,upmf,ucdf,upmfcdf,mx,ustart,LB,UB,prlevel=0)
# discretized multivariate normal/Gaussian with univariate marginal
# regression model being specified by upmf/ucdf and common
# regression parameters for each margin
MVNlatent2(ydat,xdat,nrep,unllks,upmfs,ucdfs,upmfcdfs,mx,ustart,LB,UB,prlevel=0)
# discretized multivariate normal/Gaussian with univariate marginal
# regression models being specified by upmfs/ucdfs and different
# regression parameters for each margin
vinebivEvsO1(ydat,xdat,nrep,mx,uparam,upmfcdf,cpar,A,pcop)
|
rho |
latent correlation parameter |
param |
common parameter vector for univariate margin of the latent bivariate normal model; assumed estimated in a previous step in latentBVNnllk1 |
par1 |
univariate parameter vector for univariate margin 1 of the latent bivariate normal model; assumed estimated in a previous step |
par2 |
univariate parameter vector for univariate margin 2 of the latent bivariate normal model; assumed estimated in a previous step |
uparam |
univariate parameter vector, common to all univariate margin |
ucdf |
function for univariate cdf |
upmf |
function for common univariate pmf for ieenllk |
upmfcdf |
function that efficiently computes univariate pmf and cdf up to an upper limit mx |
unllks |
vector of strings for univariate negative log-likelihoods |
upmfs |
vector of strings for pmf functions for margins 1,...,d |
ucdfs |
vector of strings for cdf functions for margins 1,...,d |
upmfcdfs |
vector of strings for pmf/cdf functions for computing up to mx |
mx |
bound used for Expected vs Observed tables in univariate/bivariate margins |
cdf1 |
function for univariate cdf for margin 1 |
cdf2 |
function for univariate cdf for margin 1 |
xdat1 |
covariates for margin 1 (nxq), q=#covariates |
xdat2 |
covariates for margin 2 (nxq), q=#covariates |
xdat |
covariate matrix ((n*d)xq), q=#covariates, d=#repeated measurements |
y1 |
count response 1, vector of length n |
y2 |
count response 2, vector of length n |
ydat |
count response, vector of length n*d |
nrep |
#repeated measurements per subject |
ustart |
starting parameter point for univariate model |
LB |
lower bound for param |
UB |
upper bound for param |
A |
dxd vine array with 1:d on diagonal |
pcop |
function for copula cdf (with scalar parameter) in tree 1 of vine |
cpar |
parameter values for pair-copulas in tree 1 of vine |
prlevel |
print.level for nlm() |
negative log-likelihood value for latentBVNnllk1, latentBVNnllk2;
negative of sum of univariate log-likelihoods for ieenllk;
list(uparam vector or parmat matrix,rhvec,E1arr,O1arr,E2arr,O2arr) [univariate parameter estimates. latent correlation estimates, Expected and Observed counts for univariate/bivariate] for MVNlatent1,MVNlatent2;
list(rhosub,E1arr,O1arr,E2arr,O2arr) for mvnbivEvsO1 where rhosub is based on subset and rhopar is based on full data set;
list(E2arr,O2arr) for vinebivEvsO1 with Expected and Observed counts for bivariate margins in first tree of vine.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 | # count regression model with common betas for all margins: longitudinal count
data(rwmsubset)
rwm=rwmsubset
rwm$agec=(rwm$age-50)/10
rwm$ageq=(rwm$agec)^2
rwm$handfra=rwm$handper/100
xdat=cbind(rwm$sex,rwm$agec,rwm$ageq,rwm$hsat,rwm$handfra,rwm$univ)
xdat=as.matrix(xdat)
ydat=rwm$docvis
nc=ncol(xdat)
nc1=nc+1
out=MVNlatent1(ydat,xdat,nrep=5,upmf=nb1pmf,ucdf=nb1cdf,upmfcdf=nb1pmfcdf,
mx=7,ustart=c(1.7,.3,.2,.1,-.2,.7,-.5,1.5),
LB=c(rep(-20,nc1),0),UB=rep(10,nc1+1),prlevel=1)
print(out$uparam)
print(out$rhvec)
#
# GP1 count regression models for each margin: uncommon regression coefficients
data(kzrepmeas)
kz=kzrepmeas
kz$agehun=kz$age/100
xdat=cbind(kz$agehun,kz$sex,kz$msmok)
ydat=kz[,6]
nrep=4
outgp1=MVNlatent2(ydat,xdat,nrep,unllks=rep("gp1nllk",4),ucdfs=rep("gp1cdf",4),
upmfcdfs=rep("gp1pmfcdf",4),
mx=3,ustart=c(0,0,0,0,1),LB=c(-20,-20,-20,-20,0),UB=rep(10,5),prlevel=0)
print(outgp1$uparam)
print(outgp1$rhvec)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.